home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19950726-19950929
/
000328_news@columbia.edu_Sat Sep 9 05:49:15 1995.msg
< prev
next >
Wrap
Internet Message Format
|
2020-01-01
|
5KB
Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA08415
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun.cc.columbia.edu>); Sat, 9 Sep 1995 15:01:54 -0400
Received: by apakabar.cc.columbia.edu id AA28248
(5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Sat, 9 Sep 1995 15:01:53 -0400
Path: news.columbia.edu!sol.ctr.columbia.edu!news.uoregon.edu!usenet.eel.ufl.edu!news.mathworks.com!newshost.marcam.com!usc!howland.reston.ans.net!news.sprintlink.net!cs.utexas.edu!news.cs.utah.edu!cc.usu.edu!jrd
From: jrd@cc.usu.edu (Joe Doupnik)
Newsgroups: comp.protocols.kermit.misc
Subject: Re: MS-DOS Kermit 3.14 Second Edition
Message-Id: <1995Sep9.114915.60947@cc.usu.edu>
Date: 9 Sep 95 11:49:15 MDT
References: <3tf3n7$2tl@apakabar.cc.columbia.edu> <42qciv$c6k@cpmt.cyberport.net>
Organization: Utah State University
Lines: 67
Apparently-To: kermit.misc@watsun.cc.columbia.edu
In article <42qciv$c6k@cpmt.cyberport.net>, Rick Owens <rowens@cyberport.net> writes:
> fdc@watsun.cc.columbia.edu (Frank da Cruz) wrote:
>
>>As announced a few weeks ago, MS-DOS Kermit 3.14 has been slightly updated
>>in several ways:
>
>> . A new KERMIT.EXE fixes a problem with TCP/IP ARP resolution.
>> . New patch files have been issued.
>> . A new font utility is now included.
>
>>The ZIP file containing these updates was placed in kermit/msdos for a
>>period of testing, and seems to have checked out OK, so it has now become
>>the official version. Effective today (July 5), the new ZIP, EXE, and PCH
>>files have been installed in all the regular places:
>
> Um, there may still be a problem with TCP/IP ARP resolution. First, some
> background: Our campus network is composed of an Ethernet backbone and an
> Arcnet backbone. The two are connected through our primary Novell file
> server (3.11, FWIW). We've been using Kermit with ODI drivers to connect
> to the Unix hosts on our network (all such hosts are connected to the
> Ethernet backbone) for about two years now. We're planning to make some
> major changes in the next few months, and in preparation I've been updating
> our set of Kermit scripts for MS-Kermit 3.14 and to take advantage of other
> changes to available software, servers, etc.
>
> Several days ago I downloaded the updated 3.14 mentioned above and today I
> put kermit.exe and the patch files on the file server for everyone to
> access. I found, though, that our users on Arcnet machines could no longer
> access our Unix hosts; Kermit would complain that it could not find the
> address of the gateway using ARP. (Ethernet connected machines work just
> fine with the same set of files.) I suspected that there might be a
> problem with the patch file so I renamed it; same result. I tried tweaking
> various settings without any luck; finally I thought to try the earlier
> version of 3.14. That fixed the problem!
>
> The machines in question are running MS-DOS 6, 5, and 3.3. The ODI driver
> is called TRXNET; the boards are mostly Compex brand. The only difference
> between a working setup and a non-working setup is which version of the
> executable and patch file is in place.
>
> I said 'may' in the first line above because I'm not very familiar with
> TCP/IP internals and how they're interacting with ODI and the Arcnet driver
> and card, and I don't have the resources here to do comparison testing. I
> would not be surprised to find that there's something peculiar in the
> TRXNET board driver which is responsible for the error message from
> MS-Kermit re. ARP. It could also be something strange with the NLM which
> forwards IP and ARP packets between media. However, in case it is a Kermit
> problem, I figured I should let everyone know about it. If anyone needs
> more information about our setup (e.g. file dates, or whatever) please ask.
----------
Ugh(tm). Arcnet with ODI has its share of problems when it comes
to ARP. An ARCnet MAC address is one byte long, yet ODI provides six byte
MAC addresses. Which end of that string will the byte appear? Undocumented.
The medium ident appearing in an ARP packet reflects the kind of wiring,
6 for Ethernet and presumably 7 for ARCnet. I did make some changes in the
MAC address extraction procedure in MSK mark II, and maybe something got
broken.
I tested earlier MSK's with ARCnet, and honestly I found that
arrangment to be flakey at best. It's not MSK but rather whatever IP
routing a NW server does in that case. I found that connections would
cease operating after a while or be reluctant to progress.
I'll see if I can resolve the situation. It's a little awkward because
I have to take apart machines and move them to construct an ARCnet environment
and then route IP to Ethernet. This may take several days because I'm eyeballs
deep in swamp creatures already and students are about to descend upon my
place en masse.
Joe D.