Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!zaphod.mps.ohio-state.edu!wuarchive!emory!sol.ctr.columbia.edu!cunixf.cc.columbia.edu!cunixb.cc.columbia.edu!mjf
From: mjf@cunixb.cc.columbia.edu (Michael J Flory)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: Faster version of Kermit?
Summary: Making kermit faster
Keywords: kermit xmodem
Message-ID: <1991Feb14.152648.10105@cunixf.cc.columbia.edu>
Date: 14 Feb 91 15:26:48 GMT
References: <1991Feb14.142237.5319@ms.uky.edu>
Sender: news@cunixf.cc.columbia.edu (The Daily News)
Reply-To: mjf@cunixb.cc.columbia.edu (Michael J Flory)
Distribution: na
Organization: Columbia University
Lines: 25

Eugene Styer writes:

>I have a situation where I want to thansfer some binary files home, but the
>network between the modem and the system uses causes xmodem to fail. I can get
>Kermit to work, but it runs only about 1/3 the speed my modem is capable of.
>So, does anyone know of a protocol that doesn't rely on a lot of special
>characters, and also gets reasonable throughput?  Send mail if you know
>about anything.

It sounds like another communication protocol won't help, as the problem
is the communication between the system (I gather you mean the remote system)
and the modem.  Maybe your remote system is allocating communication ports
between you and other users and that slows things up.  But since Kermit
(at least the non-sliding-windows flavor) has to wait for confirmation before
sending each packet, try setting the packet size to something longer than the
usual 94 bytes.  Then the whole turnaround process will have to happen less
often and things should go faster.  I think the maximum is 2000 bytes or 2K 
-- I set 2000 and it seems to give me about 2K (2048 bytes).  The only problem
might be if you have a noisy phone line -- the "cost" of resending a packet
is higher then.

Tho' I'm at Columbia, I'm *not* part of the Kermit team, so this is just
my amateur shot at the problem.  Good luck!

Michael Flory (mjf@cunixb.cc.columbia.edu)