[wplug] Starting and Stopping X More Gracefully

[Jonathan S. Billings]

On Jan 16, 2005, at 12:37 PM, Brad Hoover wrote:

> Hello Folks,
>
> I have been trying to iron out problems with x
> windows.  I would like to get 1024x768 (which I know
> both my monitor and video card can do), but right now
> I can only get 800x600, when I delete the 1024x768
> option from xorg.conf.
>
> I am using fedora core 3, which by default starts x
> during bootup.  So when I make changes to xorg.conf
> and add 1024x768 as an ok screen resolution it tries
> to display it at that setting and when it fails my
> monitor just goes on standby.  I then have to
> ctrl-alt-del and boot from a rescue disk or in single
> user mode to make another tweak it and try again

Most likely your problem is you are trying to run a resolution that the 
monitor cannot support.  You should set the HorizSync and VertRefresh 
to what your monitor can support.  The documentation provided by your 
monitor's vendor should have that information.

You don't need to boot into a rescue disk though.  Just fix your system 
so it doesn't boot into runlevel 5, which other people have already 
described.  You should *never* have the system automatically start X if 
you are trying to debug it.  I typically run my workstations at 
runlevel 3 until I'm sure X is stable, and even then, I usually just 
log in to the text-mode and run 'startx'.

Also, when X has started in an unsupported mode, that doesn't mean all 
is lost.  You don't even need to kill the X server.  Just hit 
control-alt-shift-F1 to get back to the linux text console (console 1, 
F2 will give you console 2, F3 will give you console 3, etc.)  You'll 
be able to log in and edit the X configuration, and even kill the X 
process.


Of couse, some people might not enjoy the fact that it's often easier 
to use the command line interface to configure X parameters than some 
tool that runs in X windows, but I've always found that if you depend 
on those kinds of tools, you run into the common chicken before egg 
problem.


--
Jonathan S. Billings <billings at negate.org>