Re: [CAD_CAM_EDM_DRO] LCD hack help

Jon Elson wrote:

>
> ......... I think, but I'm not sure, that X will need to be
> running on the
> local machine, too. I'm sure there is a tricky technical way to make
> the local machine
> accept X connections without having an X server running on its own
> console, but
> I'm sure I don't know how to do that.

X is a bit tricky, and the terms 'client' and 'server' are rather
backward from what seems normal. An Xserver is a device to SERVE
graphics, so X must run on the system with the graphic tube. It will get
commands from a client program to draw things on the screen, or it can
send network commands to another machine to run a tool, but the client
program does not need to be runing on a system with X installed.

So, EMC could run on a 'headless' system with the motors connected to
it, and get its commands over the net from the graphic server or X
display. The X client program does need to be compiled with the X
libraries, but the system it runs on does not need an X server or any of
the graphics, fonts, etc of the usual X workstation.

In EMC its even a bit simpler. The X gui program just sends and receives
network messages from the emc motion processes. No part of X needs to be
on the motion controller box.

Im writing this from a linux box with X runing on it. I have about 8
windows open to various servers around my company, all of them are doing
their work on the server, just doing their output display on my desktop
box.

This confusion has been around since X was started. I was on the
management committee at DEC that gave the money to MIT to develop X in
about 1985. It took a lot of people a long time to understand this
client server structure.

ron