Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!cernvax!ethz!chx400!ugun2b!ugobs!bartho
From: bartho@obs.unige.ch (PAUL BARTHOLDI)
Newsgroups: comp.sys.atari.st
Subject: Re: What Kermit/UNITERM bugs?
Message-ID: <467@obs.unige.ch>
Date: 5 Dec 89 08:52:46 GMT
References: <8912010813.AA04349@ucbvax.Berkeley.EDU>
Organization: University of Geneva, Switzerland
Lines: 36

In article <8912010813.AA04349@ucbvax.Berkeley.EDU>, 01659@AECLCR.BITNET (Greg Csullog) writes:
> One netter mentioned problems with UNITERM's Kermit. ...
> The ONLY Kermit I trust is UNITERM's so WHAT problems are there?

I can confirm that I do use UNITERM (2.0e) Kermit almost dayly in binary mode
with the following counterparts : Vax-vms Kermit32, Vax-vms CKermit, IBM vm-cms
Kermit, Sun unix Kermit and PC Kermit 2.32 .  My Atari 1040ST is connected at
19.2 Kbps to an Ethernet pad (Bridge cs/100) with XON/XOFF protocol.  The line
to the IBM goes through the vax, then through a x.25 line to a central pad that
connects me to the 3090 ...  From experience, I have set the packet length to
1024 (reduced to 96 max by CKermit ...) and use the CRC for error detection. 
Depending on various conditions I get no errors or a few retries, but never got
locked or with unreadable file (even a single error would be rejected by the
dearcing programs).  The mean transfer rate is almost 1KB/s with the Vax (780),
while it's almost 4 times slower with the IBM due to the many intermediates
nodes, but the error rate is similar.  I found the 1KB packets giving the best
throughput.  A 2KB is even better, but only if the retries are very rare. 

I never the less remember a discussion about a year ago (or more ?) between
Simon Poole and ? concerning binary transmission.  After reading all arguments,
i was convinced that UNITERM worked the right way, although the documentation
could have been interpreted differently.  The fact that I never had any
problem with the 5 different versions above confirms this 'experimentaly' 
I think (not a proof !)

          regards and Thanks to Simon if he reads this !   Paul

     ----------------------------------------------------------------
     |  Dr Paul Bartholdi             bartho@cgeuge54.bitnet        |
     |  Observatoire de Geneve        bartho@obs.unige.ch           |
     |  51, chemin des Maillettes     02284682161350::bartho (psi)  |
     |  CH-1290 Sauverny              20579::ugobs::bartho          |
     |  Switzerland                   +41 22 755 39 83       (fax)  |
     |                                +41 22 755 26 11       (tel)  |
     |                                +45 419 209 obsg ch  (telex)  |
     ----------------------------------------------------------------