Columbia Kermit/kermit.zip/newsgroups/misc.19950329-19950528/[e-mail]_Thu May 11 06:26:00 1995.msg

Internet Message Format | 2020-01-01 | 4KB

Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA08535
  (5.65c+CU/IDA-1.4.4/HLK for <[e-mail]>); Thu, 11 May 1995 02:16:52 -0400
Received: by apakabar.cc.columbia.edu id AA16835
  (5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Thu, 11 May 1995 02:16:50 -0400
Path: news.columbia.edu!sol.ctr.columbia.edu!howland.reston.ans.net!swrinde!gatech!europa.chnt.gtegsc.com!library.ucla.edu!info.ucla.edu!news.bc.net!unixg.ubc.ca!rover.ucs.ualberta.ca!tribune.usask.ca!canopus.cc.umanitoba.ca!newsflash.concordia.ca!csbn2.concordia.ca!cabilio
From: [e-mail]
Newsgroups: comp.protocols.kermit.misc
Subject: Terminal emulation in script
Date: 11 May 1995 01:26 -0500
Organization: CSBN Concordia University
Lines: 48
Distribution: world
Message-Id: <[e-mail]>
Nntp-Posting-Host: csbn2.concordia.ca
News-Software: VAX/VMS VNEWS 1.41
Apparently-To: [e-mail]

Is there no way to have MS-Kermit perform terminal emulation while a script
is running?  Here are two examples of where this would be desirable (I was
using MS-Kermit 3.14 under DOS 6.2 and VMS 5.5-2):

First, suppose you have an automated login procedure to a VMS host.  The
script sends the appropriate information in response to the Username: and
Password: prompts, then waits for the $ prompt before it resumes.  Now our
system SYLOGIN.COM issues a SET TERMINAL/INQUIRE to set the terminal
characteristics upon login, which works fine if Kermit is in CONNECT
(emulation) mode, but not if it is executing the script command INPUT 30 $.
After the INQUIRE times out, VMS issues an error message and thenceforth
treats the terminal as a dumb TTY.

One can get around the first problem by including a post-login script
command like OUTPUT SET TERMINAL/DEVICE=VT300\13, but this is a kludge.
And here is a second, more difficult example.  Suppose an MS-Kermit script
includes an OUTPUT command that invokes a VMS .COM procedure.  Among other
things, the .COM issues an escape sequence to the terminal to put it into
autoprint mode, so that some information should come out on the printer
attached to the PC.  Again, works fine if Kermit is in CONNECT mode, but if
it is executing an INPUT command at the time then the host's escape
sequence is ignored.

At first I thought I was missing something, since MS-Kermit's developers
seem to have thought of everything in this mature product.  Yet I could not
find this problem addressed in my copy of "Using MS-Dos Kermit" 2nd ed. by
Christine Gianone, nor in KERMIT.UPD, nor in the FAQ, nor in any articles
current in this newsgroup.  It is all the more surprising because one
commercial product I have used in the past includes the equivalent of an
INPUT command with terminal emulation while it waits.  Christine does note
that no terminal emulation occurs during the INPUT command, but I found
this to be equally true with the PAUSE command, leading me to infer that
terminal emulation occurs only in CONNECT mode.  Is this correct, or am I
just misunderstanding something?

There is only one way I can see of accomplishing the kind of thing I
described.  I have not yet studied the new APC feature in detail or tried
it out, but I can see how it might be used to trade control back and
forth between the host and MS-Kermit: the Kermit script starts up a .COM on
the host and goes into CONNECT mode, while the .COM does its thing and then
issues an APC to transfer control back to a Kermit script, and so forth.
Does this sound feasible?  If so, it is better than nothing, but it seems a
cumbersome solution to what appears to be a simple problem.

Much obliged for any observations you might have.

Steve Cabilio ([e-mail])