Date of last revision: 05/09/2000

It is sometimes useful to run programs on one system while you are sitting at a nother. Common reasons for doing this include accessing a different operating system, finding a more lightly-loaded system, or using an application only installed on certain systems.

This is easy to do, but it is also easy to make mistakes if you do not do it regularly.

X Windows applications use the DISPLAY variable to tell them where to send the graphical output, so the first thing you must do is determine your display name. This is generally either the hostname or IP address with :0 or :0.0 added. If you are sitting in the Design Center, the system should be labelled with its hostname. Otherwise, the command "echo $DISPLAY" should print out your display; try a simple X Windows command such as "xclock" to make sure the display is set correctly. Note: If you cannot determine your display, you will have to ask your site system administrator or internet service provider.

Type "xhost +" to allow other systems to send their display back to their system. You can also specify specific systems to be allowed access; see the man page for details.

You can now to connect to the remote system with telnet, rlogin, or ssh.

Finally, you must set the DISPLAY variable to the correct value. Sometimes it is set correctly automatically, but not always--type "echo $DISPLAY" to check. If your shell is sh, ksh, zsh, or bash (the Design Center default), the command "DISPLAY=whatever ; export DISPLAY" will set DISPLAY to whatever. For csh and tcsh users, "setenv DISPLAY whatever" should do the trick.

You are now ready to run your X Windows applications. Each time you connect to a remote system, you will have to set DISPLAY, but the xhost command is only needed once each time you log in to your local system.