> At 11:10 PM +0200 6/20/02, you wrote:
> 
> :)Yes :-)
> :)I'm still at that OF 2.0.2 prompt :_)
> :)
> :)Tried the various boot options, but can't seem to get it to boot the CD.
> :)(got the ISO from the main site, burned it with Nero)
> 
> Did you make the common mistake of putting the ISO as one file on the CD-R
> instead of using the ISO as the filesystem for the CD-R?

No.

>  What happens when you use the Open Firmware dir command to get a listing
> of the CD?

Nothing, but that could be because I was quering the SCSI controller. It
turned out that the 4400 has the cdrom connected to IDE port 1.

> :)Will try to go the floppy way now.
> 
> That should work as well.

It worked, and it is up. Some loose ends/questions.

- The comslot ethernet is detected but doesn' work,
   a RTL 8029 in a PCI slot works fine.
- I can't see any detection of an audio device.
- X starts up, but the mouse doesn't work, and neither does
	ctrl-alt-backspace. 
- The memory footprints of apps seem to be quite large, compare to i386.
- I turned autoboot on again (to allow the machine to come up again after a
crash/remote reboot). However I don't have MacOS anymore (can't be on the
same disc on pre 3 OF). Is there a bootvariables prog for NetBSD?  

I'm supping now, and hope to compile a new kernel tonight, to fix some
of these issues  
 
> :)Also, I didn't see much about partitioning. Do I have to partition it using
> :)Apple software _before_ I boot to NetBSD, like on my IIci?
> 
> Nope.  I had hoped that was clear in the INSTALL notes -- for pre-Open
> Firmware 3 models, you use the NetBSD installer to wipe and partition your
> hard drive.

The documentation is fine and extensive, but that is also its problem. It is
simply too huge if you are at the same time new to PPC hardware,
OpenFirmware, and have only very superficial knowledge of MacOS. Luckily,
I'm recently able with NetBSD and *nix itself, so once it started booting,
I installed and configured in nearly one go.

To give you an idea: I spent 2 full evenings to get it to work. And I
already tried to read in at work before.

The only way to fix this IMHO is to make model (or family) specific
summaries per model. to create a more chronological treatise of the
procedure.