On Fri, 2004-07-30 at 11:49, Jim Cheetham wrote: > Perhaps the answer is to run an X for local displays, and another X for > the remote ones? But I sort of thought that was the whole point of > Xaccess and XDMCP ...
Under X11 the server is what you are looking at. A print-server has a printer attached, a file server has a disk-drive attached, a proxy server has a network attached, an X11 server has a monitor attached. Therefor, you will be running two X11 servers: one on your normal (big) machine, and one on your terminal. (You do not have to run an X11 server on the normal (big) machine.) When using your terminal, the remote machine (the big one with hard-disk and a stonking great CPU) is the *client*. The applications you run on the client machine request the server (the tiny little box) for access to the most valuable part of any computer system: the user. Sometimes the client applications will request services that the server is unable to provide, such as using the OpenGL or X-Video extensions. In this case you will get errors, even if the client machine has the requisite libraries installed. What XDMCP does is allow the XDM *client* (such as KDM, GDM, or XDM) to attach to the X11 server and have the entire *session* run as a client. The terminal (the little box) will still need to run a X11 server so the XDMCP client can connect to it. (I have an entire presentation on this sort of thing ready for a LUG meeting, if you are interested ;) ) -- Michael JasonSmith http://www.ldots.org/
