What happened was I (accidentally) booted in 16 bit at 832x624, a combination I hadn't tried, and it's right there! It looks like this _is_ the solution for my internal video card according the somewhat "encrypted" readme of OSFA.

Just that the screen base was wrong, part of the left screen was shifted to the right. I haven't found a definitive message saying it's been fixed with generic kernel 1.3.2 but I'll definitely try it later.

Cheers,
Charlie

------
PS: read following
Re: X Server probel (with Centris650) By Ken Nakata @synap.ne.jp

On Wed, 14 Oct 1998 18:23:42 +0200 (CST), Peter Mahnke,,, wrote:
  
  The XServer starts, but the display has moved to the right side ( I believe the Base address does not
  work ). 


That is correct, and the problem is not with the X server but with the kernel (more precisely with the grf driver).

  
  I read the FAQ and the Instructions for the OSFA Server, but I got no hint how to fix this Problem -
  does anybody out there know how to fix it.


This problem is supposed to be fixed in current (however, current mac68k source has been broken for a
while and it's compilable but compiled kernels will not run). Try a current snapshot kernel from
ftp://ftp.netbsd.org/pub/NetBSD/arch/mac68k/binary/kernel/. Since Q650 uses NCR53C96 SCSI
controller rather than 53C80, ncrscsi or sbc driver is a non-issue. Try GENERIC or GENERICSBC and if
you have trouble with your ADB devices, try one of the MRG kernels.

Oh, BTW, this is definitely going to cause the "ps: proc size mismatch" problem. See the NetBSD/mac68k
FAQ for more details.

Hope it helps,

Ken
-------

"Michael R. Zucca" wrote:
> 
> >But...with 800x600 resolution, startx hangs after it prints a line telling me
> >the correct resolution and depth. It happens with both the MacX68k that comes
> >with the xbase.tgz (1.3?) and OSFA's, and regardless whether in 1 bit or
> >else.
> >The FAQ told me to give it as much time as possible, but it really seemed it
> >hung there.
> 
> Ok, you mentioned 1.3? You may want to try a later version of the kernel.
> 1.3 had problems getting the base address of the frame buffer on Quadras
> in certain resolutions.
> 
> Later kernels (1.3.2?) have a patch that reads the video controller registers
> to get the frame buffer offset.
> 
> Hope this helps!
> 
> _______________________________________________________________________
>  Michael Zucca - mrz5149@acm.org - http://www.rit.edu/~mrz5149/
>  "I will choose a path that's clear. I will choose Freewill. "
>   --Rush, Freewill
> _______________________________________________________________________