Subject: Re: Colorkit kernel boot failure/X problems
To: Colin Wood <ender@is.rice.edu>
From: None <masami@fa2.so-net.or.jp>
List: port-mac68k
Date: 11/29/1996 18:45:07
> > >Also, Ken Nakata just recently posted that a far more recent color X server has
> > >been compiled and should be available (on puma, I think ?).
> > 
> > I've tried using that X server and it won't load up for me.  Errors.  Would
> > that be the result of trying to run it without a color kernel (since they
> > don't work for me, I've only tried running the new color X server under my
> > usual GENERIC 1.2 kernel and it fails)?
> 
> Yes, this is the reason for this problem.  The color X server makes grf 
> calls that _only_ exist in the colorkit kernels.

*Ouch*, it shouldn't be the reason although I cannot say that it must
be something else.  As I wrote (somewhat vaguely?) in my previous
message, the new color X server should

	1) first attempt to use color kernel grf calls to init
	   the screen,
	2) and, if and only if 1) fails, then assume that the
	   screen is in 1-bit B&W mode, which would work on any
	   kernel including non-color kernels.

I've tried the server on both a color kernel and a non-color kernel,
and they seem to work (at least for me).

Dinsdale, could you give me the exact error messages from the server?
You can save them by redirecting both stdout and stderr of xinit or
startx to a file.

> > >This really depends on how you happen to have started X.
> > 
> > "startx".  Enter.  :)

There's more to it than that.  It depends on what programs you list in
your .xinitrc file, and generally, X exits when you terminate the
*last* program in the file.  If it's xterm, then you can exit X by
exitting from the xterm, as Colin writes below.

> Try typing "exit" in the xterm named "login".  That should be the one 
> that controls the whole process.  If not, you may need to write your own 
> .xinitrc file which sets things up a little bit better.

Later,

ken