home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19941031-19941221
/
000402_news@columbia.edu_Wed Dec 7 03:59:22 1994.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 AA21197
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun.cc.columbia.edu>); Mon, 12 Dec 1994 11:40:57 -0500
Received: by apakabar.cc.columbia.edu id AA26598
(5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Mon, 12 Dec 1994 11:40:54 -0500
Path: news.columbia.edu!spcuna!solaris.cc.vt.edu!news.mathworks.com!news.duke.edu!godot.cc.duq.edu!newsfeed.pitt.edu!gatech!swrinde!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: Flow Control in MS-DOS Kermit 3.14
Message-Id: <1994Dec7.095922.34783@cc.usu.edu>
Date: 7 Dec 94 09:59:22 MDT
References: <3c2me7$d29@sundog.tiac.net>
Organization: Utah State University
Lines: 76
Apparently-To: kermit.misc@watsun.cc.columbia.edu
In article <3c2me7$d29@sundog.tiac.net>, ciaraldi@max.tiac.net (Michael Ciaraldi) writes:
> Could I get an explanation of how flow control works in MS-DOS Kermit 3.14?
> I've read the MS-DOS Kermit book and the KERMIT.UPD file, but they
> only tell you that you can set flow control to NONE, XON/XOFF, and RTS/CTS,
> and that the XON/XOFF can be set separately for ingoing and outgoing data.
> The documentation does not give any detail beyond that.
>
> I'm especially interested in XON/XOFF.
> Here are my questions:
>
> 1) Does "ingoing" mean from the host into the serial port of the PC
> and then into Kermit, and does "outgoing" mean from Kermit out
> the serial port to the host?
Yes. INCOMING means flow control bytes to/incoming Kermit on the PC,
OUTGOING means from/outgoing from the PC to the remote host, and XON/XOFF
means in both directions.
> 2) Is flow control in Kermit supposed to work the same during
> terminal emulation as during Kermit file transfer?
> And is it the same during TRANSMIT?
Transport channels service all higher layers the same way.
Recall that flow control is a datalink affair, not high level operation.
> 3) What is the correct setting to have Kermit pass any ^S and ^Q
> characters on to the host, without further processing?
> This would emulate a terminal that implements no flow control at all.
That's a host problem, yes?
> 4) What is the correct setting to have Kermit freeze the screen when
> the user types ^S, and unfreeze it when the user types ^Q,
> but not send the ^S and ^Q to the host?
The host must stop sending, then the screen will have nothing
new to display.
> 5) What is the correct setting to have Kermit freeze the screen when
> the user types ^S, and unfreeze it when the user types ^Q,
> but also send the ^S and ^Q to the host?
> This would emulate a real VT-100, I think.
See 4.
> 6) What is the correct setting to have Kermit send a ^S to the host
> if the host is sending data faster than Kermit can process it,
> then send a ^Q when Kermit can again process new data?
> This is also the way a real VT-100 works.
> In this mode, what is the maximum number of characters the
> host can send after Kermit sends the ^S, without Kermit losing any?
That's the essence of flow control. OUTGOING or XON/XOFF does
exactly this. Kermit has a several hundred byte cushion, but modems and
host buffers and other comms storage may exceed that. Kermit does not
stop working when it sends and XOFF to the host, so overrun is a gradual
effect. A real VT100 is much slower than MSK and has only a few bytes of
cushion.
> 7) What is the correct setting to have Kermit recognize a ^S sent from
> the host, then not send any data to the host until the host
> sends a ~Q? In this mode, how many characters can the user
> type without Kermit losing any?
>
> 8) Suppose Kermit is doing a TRANSMIT without PROMPT, so that it is
> just sending the contents of a file out the serial port.
> What is the setting to have Kermit recognize a ^S from the host
> as a signal to not send any more data, then resume sending
> when the host sends a ^Q?
Again, basic flow control. The host is sending the XON/XOFF
flow control bytes. Deduction: use SET FLOW INCOMING or XON/XOFF.
> 9) Is all this documented somewhere? If so, I will certainly feel foolish
> for having missed it.
We don't offer a tutorial in basic datacomms so all these details
are not explained in the user's manual. The actions are, but not the way
you ask the questions.
> 10) If I want to see how this is implemented, where is it handled
> in the source code?
Serial comms are in file msxibm.asm.
Joe D.