Info-Kermit Digest Fri, 19 Dec 1986 Volume 5 : Number 19 Today's Topics: New Kermit Program for IBM 370 Mainframes with MVS/TSO Kermit-MPX version 2.3 Misuse of Okstate services C-Kermit and Xenix C-Kermit on Microport Unix Uploading Macintosh Binhexes with Kermit Bug in Apollo Pascal Kermit Prime kermit connect Kermit and X.25 Problem with HP9845 Kermit HCP and HC6 Versions of Kermit Need Fast De-BOOing Program for the Amiga Data General MV2000 Kermit 3.1 - An interesting speed problem Symbolics V7.0 Kermit ---------------------------------------------------------------------- Date: Thu 18 Dec 86 15:31:12-EST From: Frank da Cruz Subject: New Kermit Program for IBM 370 Mainframes with MVS/TSO Keywords: MVS/TSO Kermit The new IBM 370-series mainframe MVS/TSO Kermit from the US National Institutes of Health (NIH), announced in the previous Info-Kermit digest, is now available in the Kermit distribution areas under the prefix TSN. There are 21 files, comprising a total of about 3 megabytes, including three documentation files and a TSO help file. The program is written in "ALP", which is a preprocessor for 370 assembly language developed at NIH. The ALP preprocessor, also supplied, is written in PL/I. For those who do not have PL/I or do not wish to bother with the source programs, a hexidecimal-encoded object file is provided, along with an assembler program to decode it into a binary object file; this can be linked with a tailorable module (written in straight assembler) in which site dependencies, such as the ASCII/EBCDIC translations, are specified. Before deciding to transfer all 3 MB from Columbia over a network, first get the file TSNKER.TXT, which explains which files are which, and then only get the ones you really need. Thanks to Roger Fajman at NIH (RAF@NIHCU.BITNET) for submitting this program to us. Roger participated in the design with Dale Wright, who then did the programming. The new program has many advanced features over previous TSO Kermit versions, including server mode, binary file transfer, file interruption, 8th-bit prefixing, run-length encoding, alternate block check types, and support for both 3705-style line mode and Series/1-style full screen emulation. It is hoped that this new version will render the old University of Chicago (linemode only, circa July 1984) and University of Toronto versions (Series/1 only, March 85) obsolete. Reactions from TSO sites will be appreciated, in the interest of keeping redundant Kermit versions at a minimum. Reactions from users of the Pascal/VS version from the University of Bern (linemode only, Sept 86) will also be appreciated. For the present, the Chicago, Toronto, and Bern versions remain available in the Kermit distribution under the prefixes TSO, TSO, and TS2, respectively. For the future, there may still be another TSO Kermit program on the horizon, a result of a cooperative effort among IBM mainframe Kermit sites to develop a Kermit program that is portable among all IBM 370 mainframe operating systems (no estimate as to when this will be ready, but IBM mainframe system programmers who are interested in developments in this area may send mail to IBM-KERMIT@CU20B or IBM-KERMIT@CUVMA). ------------------------------ Date: Sun 7 Dec 86 22:40:44-MST From: Mike Niswonger Subject: Kermit-MPX version 2.3 Keywords: Gould, SEL, MPX The latest version of Kermit-MPX (GM2) for Gould machines is now available for you to FTP from my directory at Simtel20. All files are found in directory PD:GM2KERM.*. This release 2.3 of Kermit-MPX, which is also being submitted to the Gould User's group library. -- Mike Niswonger [Ed. - Thanks Mike! The files are also available through KERMSRV at CUVMA and on CU20B using FTP (user ANONYMOUS), any password. These files have replaced the old KER:GM2*.* and the old files are now in KO:GM2*.*] ------------------------------ Date: Mon, 08 Dec 86 18:50:27 -0600 From: Mark Vasoll Subject: Misuse of Okstate services Keywords: Okstate, Oklahoma State Fellow Kermiters, It saddens me to inform you all that various people have started to abuse the UUCP login "uucpker" on Okstate by attempting to use us as a jumping off point for mail. While we enjoy providing the home for Kermit access via UUCP, we have no desire to become a major mail relay point for various random systems. To this end, I have had to place a restriction on our UUCP to not accept the "rmail" command from any system using the uucpker account. This will, unfortunately, make it more difficult for people to report problems to okstate!uucp-support, but there is very little I can do about it. Please use the following mail paths when reporting problems with the Kermit distribution (no phone calls please). Internet: uucp-support@a.cs.okstate.edu or: uucp-support%a.cs.okstate.edu@csnet-relay.arpa UUCP : {cbatt, cbosgd, ihnp4, pesnta, rutgers, seismo}!okstate!uucp-support Mark Vasoll Computing and Information Sciences Internet: vasoll@a.cs.okstate.edu Oklahoma State University UUCP: {cbatt, cbosgd, ihnp4, pesnta, Stillwater, Oklahoma rutgers, seismo}!okstate!vasoll ------------------------------ Date: Tue, 9 Dec 86 08:05:48 EST From: Marshall_DeBerry@um.cc.umich.edu Subject: C-Kermit and Xenix Keywords: Xenix, C-Kermit RE: Info-Kermit Digest V5 #18 I recently put the latest C Kermit (4D(061)) up on a Tandy 16/6000 running Xenix 3.1 with no problems. I used the make sys5 command to compile; only changes needed were to include /sys/typedefs.h for the void identifier found in one file. Everything seems to be working ok so far. ------------------------------ Date: Mon, 15 Dec 86 10:17:15 EST From: es!Dan_Senie%anvil.UUCP@harvard.HARVARD.EDU Subject: C-Kermit on Microport Unix Keywords: C-Kermit I just compiled 4D(061) using 'make sys3'. It work right from the start. Many people have sent in messages about trouble getting Kermit to run on XENIX. I had used XENIX for a time, but found the include files and compiler contained many errors and incompatibilities with real system V Unix. Microport Unix is a full System V for the IBM-AT. Unlike XENIX, it is based on the current Sys V sources and therefore Kermit was easy to get running. Those of you using XENIX may want to investigate this new package. Dan Senie PS. I am not connected in any way with Microport Systems, ATT, etc. UUCP: ...!harvard!anvil!es!Dan_Senie (case is important!) ARPA: anvil!es!Dan_Senie@harvard.harvard.edu ------------------------------ Date: Sun, 7 Dec 86 11:29:35 pst From: gould9!joel@nosc.ARPA (Joel West @ Western Software Technology) Subject: Uploading Macintosh Binhexes with Kermit Keywords: UNIX, MacKermit, hqx I've been meaning to send this along for a while, but I wasn't sure anyone was interested. A query on INFO-MAC prompted this posting. It is a UNIX csh script (MACKERM) that I use to upload binhexes using kermit. If for a file foo.hqx, it first sends the header (before the binhex) as 'About foo', then sends foo.hqx. I use the header to later note the origin on the final unhexed file. To make this work, you need to take the first file (usually a short About... file) manually, selecting the target directory. Then place your MAC in server mode, since each 'kermit' command at the UNIX side is a separate session, and the Mac would otherwise quit file transfer mode. Oh yeah, it also works on ordinary files, preserves actual names ('FooBar' instead of 'FOOBAR') and NEVER transfers a directory. So it's real useful to type mackerm * place the Mac in server mode, and walk away. Joel West joel%gould9.UUCP@NOSC.MIL ihnp4!gould9!joel ==File ===== #!/bin/csh # by Joel West, ihnp4!gould9!joel, 11/14/86 # Script to send binhexes and other files # Also preserves exact upper/lower case name set noglob foreach f ($*) if (-d $f) continue set typ=$f:e switch ($typ) case hqx: case hex: set r=$f:r set T1=/tmp/kermdat_$$ set T2=/tmp/kermhqx_$$ rm -f $T1 $T2 cat $f | cutat '(This' $T1 $T2 kermit -s $T1 -a "About $r" kermit -s $T2 -a $f rm -f $T1 $T2 breaksw default: kermit -s $f -a $f endsw end == EOF == [Ed. - Thanks. This hint has been added to the Macintosh Kermit .BWR file, CKMKER.BWR.] ------------------------------ Date: Fri 14 Nov 86 10:34:23-PST From: Ted Shapin Subject: Bug in Apollo Pascal Kermit Keywords: Apollo Kermit We found that the Pascal kermit for the Apollo inserts extra line feeds every 256 characters if there is no line feed carriage return in the file. We traced it to the I/O in the Pascal program which can only read 256 byte lines. [Ed. - Thanks for the information. This message is now in the file KER:APOLLO.BWR.] ------------------------------ Date: Wed 10 Dec 86 10:45:54-PST From: Bob Larson Subject: Prime kermit connect Keywords: Prime Kermit I have a partially working one, based on the old version of prime kermit. I'm not even sure if I have a complete set of sources for it on tape any more. Known bugs include dropping characters in connect mode, problems with some of the message strings (incorect use of ioa$), and not supporting 2400 baud. (Writen before prime standardized the "optional" speeds at 2400 and 4800 baud for autobaud.) I have been considering making improvements to the current prime kermit, with support for amlc lines among them. ------------------------------ Date: WED, 10 DEC 86 17:17:13 BST From: CJK @ UK.AC.SALFORD.R-D Subject: Kermit and X.25 Keywords: C-Kermit, X.25 Points from infodigest 5/17. First Steve Walton & Telenet: We over this side of the pond are quite used to using Kermit over X.25 networks (posing as an X.29 terminal). It works fine, at least in the micro-mainframe mode, with call originated by the micro. Efficiency should not be affected much until the ratio: time-through-network / front-end-delay reaches about 5. The crucial thing is to set up the X.3 parameters in the PAD so that the whole Kermit packet gets into a single X.25 packet; it is bound to be shorter than the 128-byte maximum on X.25. In general you have to use a transparent-type mode to let the SOH get through, but there is no need at all to forward on every character (which simply fills up the X.25 window with single-character X.25 packets). Provided the EoL is a forwarding character, which it will be if it is CR, all works fine. What you actually do depends on the user interface provided by your PAD. This is in fact all covered in your book. Of course, network congestion can slow things down. Our X.25s, the public PSS and private JANET, have maximum time-through-networks of about 0.1 sec in all normal circumstances, so no problem. But I heard last week in a different context that the congestion on ACCUNET was so severe that holding transatlantic calls open was very difficult. If Telenet is in the same state, then you've got problems. But blame the sizing of the network, not X.25 or Kermit. Does CKermit really limit you to 65% of line-speed? I tested homegrown Kermits here (at up to about 30Kbaud) and found that, with data-compression on, I always got better than 100% of linespeed. If CKermit is really that slow, is it losing time somewhere that it needn't? Incidentally, switching over to Kuryan Thomas' problems with his autodialler: Many UARTs will not receive if DCD is not high (it's a protection against line-noise). If you have one of these, and an autodialler which does not hold up DCD while it is trying to talk to you, then you have no option but to strap DCD to one of the other pins, e.g. DSR. Chris Kennington. ------------------------------ Date: 11-DEC-1986 09:47:23 From: DGM1@UK.AC.YORK.VAXA Subject: Problem with HP9845 Kermit Keywords: HP Kermit Cross-Ref: Hewlett-Packard Kermit It's not a bug: it's a problem regarding some assumptions regarding the machine configuration, which I should have spelt out when I sent it in. We have explained/apologised to one or two people who contacted us directly, and if in a fit of absent mindedness I had not deleted a note I had from Chris regarding the configuration assumptions, I would forward it to you for a BWR file. As I mentioned when I sent it in, it does not pretend to be a full kermit, more a minmalist bulk data transfer utility conforming to the kermit protocol. It is probably possible to bypass the need for specific ROMs and so on by the use of inline assembler, however, if the ROMs are available, it is better to use the facilities provided by the HP virtual communication device. I'll get the precise details from Chris as soon as possible for you for inclusion in a BWR Sorry about the hassle, Doug ------------------------------ Date: 11 Dec 86 13:42:00 EST From: John Stewart Subject: HCP and HC6 Versions of Kermit Keywords: Honeywell Kermit I was looking at the versions file and noticed that two versions of Kermit are listed for our machine, HCP and HC6. HCP is a Pascal implementation of Kermit that was ported to the Honeywell Cp-6 operating system by Bucknell University. It is a very minimal implementation (it doesn't even do repeat counts, let alone advanced features like server mode). HC6 is much more complete then HCP, and both the original developer and myself are continuing to enhance it. I suggest that distribution of the HCP version of kermit be discontinued. Regards.. jas [Ed. - Thanks, this seems to be the concensus of opinion; the HCP version was indeed retired to Tape C.] ------------------------------ Date: Wed, 10 Dec 86 10:33 EST From: "John G. Ata" Subject: Need Fast De-BOOing Program for the Amiga Keywords: Amiga Kermit Can somebody out there who has a C compiler on the Amiga please compile CKIBOO.C and then make a BOO file out of it and send it in to Kermit Distribution as, say, CKIBOO.BOO, so that those of us without a C compiler only have to run the Basic de-booer once, on CKIBOO.BOO, and then can subsequently use the compiled de-booer which should be much faster. It takes about an hour to de-boo Amiga Kermit with the Basic de-booer. John G. Ata ------------------------------ Date: Tue, 16 Dec 86 16:25:14 est From: munnari!latcs1.oz!philip@seismo.CSS.GOV (Philip Lee) Subject: Data General MV2000 Keywords: Data General Has anyone sucessfully ported the unix kermit to Data General MV2000 running DG/UX ver. 3.00? We've tried recompiling ver 4D on it. It compiled without complaining, but gave link error. However it sort of work. We could do ascii file transfer, but any compressed file or any of those using all 8 bits would fail. The DG kermit would accept set file type binary quite happily and indicate the same, but on file transfer it would just time out and fail. In generating the DG/UX kermit we used the following: make wermit \ "CFLAGS = -D_file=_fd -DUXIII -DDEBUG -DVER2 -DTLOG -O4" "LNKFLAGS =" Could this be also related to the problem of 8 bit transfers between FIDOnet Kermit and the Pyramid? [Ed. - The FIDO question has still not been totally solved, mostly because of lack of time, but it might help you to know there's a version of Kermit specifically tailored for DG MV/UX (is that the same thing?), based on an old release of Unix Kermit, available in the Kermit distribution under the prefix DGM. There's also a C-Kermit implementation for AOS/VS on the way.] ------------------------------ Date: Sun, 14 Dec 86 14:56:29 +0200 From: Sam Gamoran Subject: Kermit 3.1 - An interesting speed problem Keywords: CMS Kermit, PROCOMM We just installed CMS Kermit 3.1 - appears to solve our problems with 2.1 particularly the program hanging and binary transfers. We use CMS Kermit through a 7171 against PCs running either MS-DOS Kermit 2.28 or Procomm 2.3 with the Kermit file transfer option. Works great with MS-DOS Kermit 2.28 but file transfer crawls with Procomm 2.3. Every packet sent by CMS-KERMIT 3.1 ends with an X-ON character. The old CMS-KERMIT 2.1 did not do this. It appears that this X-ON character causes Procomm to delay a second or so before sending the next packet. We saw on a line monitor packets going out from PC to mainframe and an instantaneous response followed by a delay. With the old CMS-KERMIT, where the main difference between packets is the absence of the final X-ON character no delay occured and transfer throughput was 400% faster! Do you know of anyone having experience with PC->Mainframe file transfer using PROCOMM who has encountered similar problems? I am also sending this query to the procomm folks. Thanks, Sam Gamoran ------------------------------ Date: Tue, 16 Dec 86 15:39 CST From: "David S. Cargo" Subject: Symbolics V7.0 Kermit Keywords: Symbolics Kermit Symbolics is upgrading their system software to version 7.0. The old Symbolics Kermit no longer works with the new system revision. Has anybody upgraded the existing version to run on the new system? I'm trying to avoid some wheel reinvention. ------------------------------ End of Info-Kermit Digest *************************