Hi Paul,

> Hi Raz,
> 
> Thanks very much for your very insightful explaination.

You are most welcome.

> I changed the /m200/ .login file so that the DISPLAY parameter sets to 
> xia:10.0. But I still got the same error message. What would be an 
> appropriate way to do this?

The sshd (which launches the shell which launches your menu which
launches xterm...) selects and sets the DISPLAY variable by itself,
typically to localhost:10.0, but it could set it to almost anything.
What is important is not what it is actually set to (it could even
change from login to login) but that you let it get passed through to
xterm without your script(s) modifying it.

Repeating for clarity: do not, under any circumstances, set DISPLAY to
anything, just let the value set by the sshd pass through to xterm
unmodified.

(The problems with setting it to xia:10.0 are that:

- once again, you are telling the xterm to connection to xia/TCP/6010,
which attempts to bypass the SSH tunneling and will fail if xia's
X-server doesn't accept remote connections (e.g. if it's a Debian box)

- it is most unlikely that xia has :10.0 display.)

> M200 is an instrument used by chemists to work out how atoms are 
> arranged in space. I would like to remotely run it from my office 
> computer. I used to do with Telnet but University strongly discouraged 
> it. They suggested the SSH X forwarding to me.

Indeed, ssh and X forwarding is a somewhat cleaner approach.

Given that you're getting snagged on your scripts, perhaps just bypass
them completely:

  xia:~$ ssh -X m200 xterm

If this doesn't work, then let's see the output from "ssh -v -X m200
xterm". Also, check that m200:/etc/ssh/sshd_config has "X11Forwarding
yes" in it (presumably it does if the machine's administrators are
suggesting that you tunnel X11 over ssh, but you never know).

- Raz

Reply via email to