^C during telnet from Win95 broken
Michael Ryan
When I telnet to a SCO OSR504 system from Win95 and hit ^C for
the interrupt key, it doesn't work correctly. I need to send another
character before processing of the ^C occurs. My stty settings
are fine. Any ideas? Any fixes? TIA.
Mike
<mi...@NetworX.ie>
---
Michael Ryan
When I telnet to a SCO OSR504 system from a Win95 client,
the ^C key doesn't function properly as an interrupt key.
Does anybody know a solution for this?
TIA
Mike
<mi...@NetworX.ie>
lenlutz
Michael Ryan wrote:
>
> Hi people,
>
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Most of what I have learned has been taught to me by others.
I guess that means I have to teach others what I have learned
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
lenlutz
and it never responded to ^c.... Have you tried the "BREAK" key ?????
Kevin W. Reed
>thats funny. i've only been usoing sco for around 9 years
>and it never responded to ^c.... Have you tried the "BREAK" key ?????
You are assuming that he has not changed the default intr to be ^C...
His message seems to indicate that he has since he says "I need to
send another character before processing of the ^C occurs." If he
still had the default intr setting this statement would not be true
since only pressing the DEL key would take action, not another
character.
Since he didn't provide the output of his stty, it is hard to know
for sure however...
It is possible that the ^C is being trapped by the host machine he is
telnetting in from or that he has other problems...
>Michael Ryan wrote:
>>
>> Hi people,
>>
>> When I telnet to a SCO OSR504 system from Win95 and hit ^C for
>> the interrupt key, it doesn't work correctly. I need to send another
>> character before processing of the ^C occurs. My stty settings
>> are fine. Any ideas? Any fixes? TIA.
--
Kevin W. Reed - Voice 602-469-5106 TNET Services - MAILBOT.COM
Mesa, Arizona U.S.A. Software Development
mailto://kr...@tnet.com - http://www.tnet.com - http://www.mailbot.com
[pgpk -a|pgp -ka] http://pgp.mailbot.com/kr...@telesys.tnet.com
PGP Fingerprint: 5F CD 01 70 C3 A3 6C 02 B7 7E CD 62 51 65 09 9B
Nikolaus Volejnik
the
the windows-telnet isn´t as well as teerminal emulations like TUN Emul. We
are using
TUN Emul for years and never had any problems with escape-sequences like ^C
or
others. Maybe this will help.
nik
lenlutz wrote:
> Thats sometime DELETE
>
> Michael Ryan wrote:
> >
> > Hi people,
> >
> > When I telnet to a SCO OSR504 system from Win95 and hit ^C for
> > the interrupt key, it doesn't work correctly. I need to send another
> > character before processing of the ^C occurs. My stty settings
> > are fine. Any ideas? Any fixes? TIA.
> >
Rob Hamilton
The windows telnet app is maybe not the best thing to use.
I dont use it myself, but I had a quick look... and funny things are
happening.
It's not a terminal emulation thing, but a telnet protocol
compatibility thing.
Now, instead of the prompt returing after hitting break or ctrl-c (whatever
your interrupt character is) you may see a single "funny looking"
character instead?
This turns out to be 242 (telnetd DM code)
which is usually sent by SCO 5.4 telnetd
in the sequence 255/242 (IAC-DM) on a keyboard interrupt.
My guess was that windows telnet wasnt processing this as expected.
So I tried running "script t" on SCO. This turns off the IAC-DM send
Guess what... windows telnet now behaved properly.
(btw that *isnt* a fix! script will create a huge log file you
probably dont want!)
Also telneting to an old box (SCO 3.2 which doesnt send IAC.DM)
was ok too.
er... At this I gave up. Getting another telnet for windows is probably
the simplest thing to do. (its pretty slow and awful anyway).
If you absolutely must use it you could try replacing the SCO telnetd with
an older version which might be compatible, but I wouldnt recommend that.
regards,
Rob
In article <36DFACB5...@zutritt.at>,
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Michael Ryan
wrote:
>Hi,
>The windows telnet app is maybe not the best thing to use.
>I dont use it myself, but I had a quick look... and funny things are
>happening.
Sorry for not explaining things more clearly in my original post.
Firstly, I have set interrupt to ^C, so hitting DEL/BREAK wasn't
the answer.
I'm using CRT as the emulator. Other emulators have been tried
by my colleagues and they all exhibit the same behaviour.
>It's not a terminal emulation thing, but a telnet protocol
>compatibility thing.
Thanks for pointing this out, Rob -- you're dead right. It's an
incompatibility between SCO OSR504 and Win95. We're actually
running an Informix application on the SCO server. We also have
a copy of the application on a Solarix x86 2.5.1 server and there is
no problem.
>Now, instead of the prompt returing after hitting break or ctrl-c (whatever
>your interrupt character is) you may see a single "funny looking"
>character instead?
Yes, sorta -- because we're running an Informix application, this
isn't visible but, you're right, the ^C just isn't being processed
correctly.
>This turns out to be 242 (telnetd DM code)
>which is usually sent by SCO 5.4 telnetd
>in the sequence 255/242 (IAC-DM) on a keyboard interrupt.
Spot on again. We submitted a telnet trace to the author of
CRT, our terminal emulator, and they pointed this out to us.
>My guess was that windows telnet wasnt processing this as expected.
Because the problem occurs with more than one emulator, it's possible
that the problem may be in the Win95 TCP/IP stack.
>So I tried running "script t" on SCO. This turns off the IAC-DM send
>Guess what... windows telnet now behaved properly.
Wow! Must try this...
>(btw that *isnt* a fix! script will create a huge log file you
>probably dont want!)
What about if I do "script /dev/null"? Anyway, this isn't a
fix in any case. I need a real solution, but it's interesting
to know this.
>Also telneting to an old box (SCO 3.2 which doesnt send IAC.DM)
>was ok too.
Interesting again.
>
>er... At this I gave up. Getting another telnet for windows is probably
>the simplest thing to do. (its pretty slow and awful anyway).
Agreed; we're already using another emulator.
>If you absolutely must use it you could try replacing the SCO telnetd with
>an older version which might be compatible, but I wouldnt recommend that.
This is the track I'm on. I want to compile the FreeBSD telnetd but,
guess what, I don't have a C compiler. I downloaded gcc 2.7.2.1
from www.sco.com but there are no installation instructions. It's
just a 'naked' tarball of the GNU stuff. I extracted it into
/usr/local, because that's where it seemed to want to go, but
it wouldn't work: it couldn't find any of the standard header files.
I located copies of stdio.h etc. under the GNU branch but these
gave parse errors during compilation.
Anybody know where to get a gcc that'll work?
TIA.
Mike
<mi...@NetworX.ie>
Jeffrey Altman
: On Fri, 05 Mar 1999 17:21:44 GMT, Rob Hamilton <ro...@my-dejanews.com>
: wrote:
: I'm using CRT as the emulator. Other emulators have been tried
: by my colleagues and they all exhibit the same behaviour.
:
: >It's not a terminal emulation thing, but a telnet protocol
: >compatibility thing.
:
: Thanks for pointing this out, Rob -- you're dead right. It's an
: incompatibility between SCO OSR504 and Win95. We're actually
: running an Informix application on the SCO server. We also have
: a copy of the application on a Solarix x86 2.5.1 server and there is
: no problem.
The problem is that the majority of telnet client authors do not
understand the subtle workings of the telnet protocol. When you send
^C to your host it sends a Telnet Flush command as out of band data.
A Telnet Data Mark command is sent to indicate where the telnet should
start processing data again.
Kermit 95 is one Telnet Client for Windows that handles this situation
correctly. See http://www.kermit-project.org/k95.html for details.
One of Kermit 95's features is a Telnet Debug mode which would help
you track down problems of this sort in a matter of seconds.
Jeffrey Altman * Sr.Software Designer * Kermit-95 for Win32 and OS/2
The Kermit Project * Columbia University
612 West 115th St #716 * New York, NY * 10025
http://www.kermit-project.org/k95.html * kermit-...@kermit-project.org
Michael Ryan
Altman) wrote:
>Kermit 95 is one Telnet Client for Windows that handles this situation
>correctly. See http://www.kermit-project.org/k95.html for details.
Thanks, Jeffrey, I'll check it out.
Bye,
Mike
<mi...@NetworX.ie>