[comp.dcom.modems] uucp over telnet connection

larry@nstar.uucp (Larry Snyder) (10/24/90)

I am having problems getting a good uucp session over
a telnet transfer - here are the details -

I call into a local machine, telnet to another machine,
then login and initiate a uucp transfer with this remote
machine.

The uucp transfer is aborted with multiple alarms and "bad header
messages".   Any ideas?  Is it even possible to use uucp over
a telnet connection?  

-- 
       Larry Snyder, Northern Star Communications, Notre Dame, IN USA 
 {larry@nstar, uunet!sco!romed!nstar!larry, nstar%larry@iuvax.cs.indiana.edu}
                     backbone usenet newsfeeds available
         Public Access Unix Site (219) 289-0282 (5 high speed lines)

grr@cbmvax.commodore.com (George Robbins) (10/25/90)

In article <1990Oct24.115921.22397@nstar.uucp> larry@nstar.uucp (Larry Snyder) writes:
> 
> I call into a local machine, telnet to another machine,
> then login and initiate a uucp transfer with this remote
> machine.
> 
> The uucp transfer is aborted with multiple alarms and "bad header
> messages".   Any ideas?  Is it even possible to use uucp over
> a telnet connection?  

Well, uucp (normally) requires an 8-bit transparent path and telnet
has a single character link kill, plus various other escapes...

> 
> -- 
>        Larry Snyder, Northern Star Communications, Notre Dame, IN USA 
>  {larry@nstar, uunet!sco!romed!nstar!larry, nstar%larry@iuvax.cs.indiana.edu}
>                      backbone usenet newsfeeds available
>          Public Access Unix Site (219) 289-0282 (5 high speed lines)


-- 
George Robbins - now working for,     uucp:   {uunet|pyramid|rutgers}!cbmvax!grr
but no way officially representing:   domain: grr@cbmvax.commodore.com
Commodore, Engineering Department     phone:  215-431-9349 (only by moonlite)

cec@cup.portal.com (Cerafin E Castillo) (10/25/90)

	Larry,
		UUCP over TCP/IP is in need of UUCP protocol 't'.
	I am not certain of the specifics on the 't' protocol, but
	when using a psuedo tty via telnet to do uucp sessions,
	the IP datagrams and delays do get in the way.

		Someone else in this group may know more and wish
	to comment.  I have seen some write-ups on uucp protocol
	't' in the AT&T SysVR3.2.2 manuals.  Good Luck!

	P.S.

	I hear that some of this is not true when telneting from a
	TCP/IP Terminal Server device (ie Xylogics Annex II/IIe).

===============================================================================
Cerafin E. Castillo                       ||      //\\  ||\\  ||
Network Consultant                        ||     //__\\ || \\ ||  Los Altos
Los Altos Networks                        ||    // ---\\||  \\||  Networks
340 Second St. #6                         ||___//      \||   \\|
Los Altos, CA  94022
(415) 941-8031      UUCP:     {apple,sun,uunet}!portal!cup.portal.com!cec
                 NTERNET:     cec@cup.portal.com

                      "...No hay mal que por bien no venga..."
===============================================================================

zeeff@b-tech.ann-arbor.mi.us (Jon Zeeff) (10/25/90)

>The uucp transfer is aborted with multiple alarms and "bad header
>messages".   Any ideas?  Is it even possible to use uucp over
>a telnet connection?  

My experience has been that newer versions of telnetd have a binary 
option that allows them to pass all characters.  With /usr/ucb/telnet,
you are going to have to do something about the escape character ^].
I've generally had better luck with rlogin -8.

-- 
Jon Zeeff (NIC handle JZ)	 zeeff@b-tech.ann-arbor.mi.us

larry@nstar.uucp (Larry Snyder) (10/26/90)

grr@cbmvax.commodore.com (George Robbins) writes:

>Well, uucp (normally) requires an 8-bit transparent path and telnet
>has a single character link kill, plus various other escapes...

right you are - as someone pointed out - use rlogin with a -8 which
works like a champ --


-- 
       Larry Snyder, Northern Star Communications, Notre Dame, IN USA 
 {larry@nstar, uunet!sco!romed!nstar!larry, nstar%larry@iuvax.cs.indiana.edu}
                     backbone usenet newsfeeds available
         Public Access Unix Site (219) 289-0282 (5 high speed lines)

billd@fps.com (Bill Davidson) (10/28/90)

In article <35219@cup.portal.com> cec@cup.portal.com (Cerafin E Castillo) writes:
[explanation of problems with telnet vs. uucp and using the 't' protocol]
>I hear that some of this is not true when telneting from a
>TCP/IP Terminal Server device (ie Xylogics Annex II/IIe).

I have a IIe on evaluation right now.  Xylogics provides a program
called "rtelnet" (reverse telnet) which is normally run at boot time.
It takes control of a pseudo tty, and puts itself in the background.
Any reads or writes to that pseudo-tty go through rtelnet to the Annex
IIe to a port specified when you invoked rtelnet.  It works great with
binary kermit transfers.  I'm having a little trouble with uucp so far
though (lot's of alarms and timeouts at 19200).  They sent me the wrong
cables so I don't have RTS/CTS flow control.  They are supposed to be
sending me the right ones soon so it may work fine when I get those and
use flow control (They are a bit hard to make since they go from a
50-pin connector to 8 DB25's).  Also, at least on my systems, if you
reset a port which has rtelent running on it, it panics the machine
which is running the rtelnet in the kernel pty driver.  I haven't had a
lot of time to explore this yet.  Has anyone else come across this?

--Bill Davidson

njs@scifi.UUCP (Nicholas J. Simicich) (10/29/90)

In article <35219@cup.portal.com> cec@cup.portal.com (Cerafin E Castillo) writes:

>		UUCP over TCP/IP is in need of UUCP protocol 't'.
>	I am not certain of the specifics on the 't' protocol, but
>	when using a psuedo tty via telnet to do uucp sessions,
>	the IP datagrams and delays do get in the way.

Actually, my measurements have found those delays to be fairly
insignificant.  I frequently uucp from my home machine to a machine at
work by calling an Annex II terminal server, and then using a fairly
complex chat script where I turn off all of the telnet options which
would get in the way.  I also telnet to port 540 on the work machine,
which is the port listened to by the uucp daemon at that end.  Modems
are Telebit T2500's, no flow control is used, and everything works
fine.   The newer releases of the Annex II software are capable of
making a binary character-by-character path using telnet to data
ports. 

I don't think that IP gets in the way, I think it is the "telnet"
protocol, which has a character escape, might convert cr to cr-nl, and
so forth.  

The uucp protocol I can use, g or t, is selected during negotiation.
You don't have to use t, and I don't.

>
>		Someone else in this group may know more and wish
>	to comment.  I have seen some write-ups on uucp protocol
>	't' in the AT&T SysVR3.2.2 manuals.  Good Luck!
>
>	P.S.
>
>	I hear that some of this is not true when telneting from a
>	TCP/IP Terminal Server device (ie Xylogics Annex II/IIe).

When we were using an RT as a terminal server, I did the same thing,
but I used a simple program which connected to uucpd on the remote
machine via tcp/ip.   All that was required was that the program send
the characters it reads from the modem to the remote end, and that the
characters read from the socket be sent to the terminal, without
changing them in any way.


-- 
Nick Simicich - uunet!bywater!scifi!njs - njs@ibm.com
SSI #AOWI 3958, HSA 318