home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
kermit.columbia.edu
/
kermit.columbia.edu.tar
/
kermit.columbia.edu
/
newsgroups
/
misc.19950929-19951130
/
000394_news@columbia.edu_Tue Nov 14 11:21:06 1995.msg
< prev
next >
Wrap
Internet Message Format
|
1995-12-25
|
3KB
Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA21170
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun>); Wed, 15 Nov 1995 08:29:54 -0500
Received: (from news@localhost) by apakabar.cc.columbia.edu (8.6.12/8.6.12) id IAA17735 for kermit.misc@watsun; Wed, 15 Nov 1995 08:29:52 -0500
Path: news.columbia.edu!sol.ctr.columbia.edu!spool.mu.edu!agate!newsxfer.itd.umich.edu!tank.news.pipex.net!pipex!news.mathworks.com!newsfeed.internetmci.com!inet-nntp-gw-1.us.oracle.com!news.caldera.com!news.cc.utah.edu!news.cs.utah.edu!cc.usu.edu!jrd
From: jrd@cc.usu.edu (Joe Doupnik)
Newsgroups: comp.protocols.kermit.misc
Subject: Re: Flow control xon/xoff <-> RTS/CTS?
Message-Id: <1995Nov14.172106.66884@cc.usu.edu>
Date: 14 Nov 95 17:21:06 MDT
References: <488ih7$jsg@pheidippides.axion.bt.co.uk>
Organization: Utah State University
Lines: 38
Apparently-To: kermit.misc@watsun.cc.columbia.edu
In article <488ih7$jsg@pheidippides.axion.bt.co.uk>, stevef@aom.bt.co.uk (Steve Fosdick) writes:
> We have a modem link, using modems with NMP5 which therefore implement
> flow control between the modem and the computer.
>
> At one end, the flow control mechanism is RTS/CTS and at the other end
> the flow control mechanism is Xon/Xoff.
>
> With this setup, file transfers sometimes get hung, and whenever this
> occurs, if I ^C from the file transfer, connected to the other end,
> and send an XON character I get a large number of queued packets in
> response.
>
> Never having had this problem when using Xon/Xoff at both ends I
> assume that kermit includes sending an Xon to the modem as part of
> it's timeout recovery mechanism in case the modem received an Xoff
> which kermit didn't send it, but doesn't do that if flow is set to
> RTS/CTS.
>
> Question: is there any detrimental effect to be had from trying the
> Xon heuristic in RTS/CTS mode?
>
> --
> Steve Fosdick Internet: stevef@aom.bt.co.uk
> Voice: +44 1473 642987 BT WEB: FOSDICK S J
> Fax: +44 1473 644607 BOAT: FOSDICSJ
> Snail: Room 210, B67, BT Labs, Martlesham Heath, Ipswich, IP5 7RE, England.
---------
What's needed is for each leg of the comms pathway to flow control
between the boxes terminating that leg, or do only end to end flow control
with transparency along the link. The latter approach is simpler but suffers
from major worries of overrunning one of the intermediate boxes and of the
overall byte storage of the comms channel (and it can be large). So Plan A
is the desirable approach.
Kermit won't do both XON/XOFF and RTS/CTS flow control at the same
time, nor should it. You will need to have a little talk with each modem
and the remote host about which kind of flow control is to be used where
(hopefully hardware RTS/CTS so the channel remains totally transparent).
Joe D.