At 6:24 PM -0500 10/26/00, Bob Nestor wrote:
>Michael Wolfson wrote:
>
> >At 2:26 PM -0700 10/26/00, Henry B. Hotz wrote:
> >:)What follows are nits.  As Bob Nestor says we need to have a simple
> >:)set of instructions that will *work* before we can start exploring
> >:)all the special cases.
> >
> >Unfortunately, after trying that, I found that OF has more exceptions than
> >rules (kinda like the English language).  Instead, I broke it down into
> >instructions for each different classes of OF (1.0.5 and 2.0.x, 1.1.22,
> >2.4, and 3) as within each class, most everything is the same.
> >
>What I thought I'd said was that we should work out the details on
>"standard" machine configurations first, then deal with the special cases
>of machines that have been upgraded with different disks, interfaces and
>such.  Since Apple has released standard systems with various versions of
>OF that don't all work the same, I think this is an excellent approach.

Memory fuzz.  What he said. ;-)

Somehow or other we need to get to a set of instructions that are 
guaranteed to get anyone started, even if they say to unplug all 
peripherals except Ethernet and one Apple-brand disk.

>My original statements about special case machines was based on my onw
>experience and belief that changing the machine configuration _may_
>introduce additional problems or mask what is really happening in OF.  I
>certainly beleive that I've seen this first-hand with my B&W G3; adding
>an IDE disk to the secondary controller created a situation where the
>CDROM became unbootable, even for MacOS.  Removing the disk cleared up
>the problem.  I have a feeling that if this is happening on other's
>machines as well it make be confusing the main issue of learning how to
>use OF to boot NetBSD from disk or CD.


> >:)My own experience is that the machine was fine without System Disk
> >:)and wouldn't boot properly after I ran it.
> >
> >Idunno, considering the number of models that are completely broken until
> >System Disk patches them, I'd rather encourage use of it.  Then, if there
> >are problems, they can just disable use-nvramrc?.  At least that way,
> >System Disk figures out the screen device, sets up load-base, and has a
> >non-threatening GUI.
> >
>I had the same experience with System Disk that Henry describes.  After
>that I'm kind of leary about using it again since I have no idea what it
>will modify.  I'd feel much more comfortable dealing with a set of
>instructions that tell me how to get into OF and set variables by hand
>and why I'm doing it.  That way I can keep a record of what they were
>before I changed them and have some hope of getting my system back if
>things don't work out.

I wonder if the systems where SystemDisk is dangerous also break down 
nicely by OF revision.  I have only used it with NetBSD on OF 1.0.5 
machines and I had the problem that drove me to building a custom 
nvramrc (fun when you don't know didley about Forth).  I have used it 
on my G4 with a useless Darwin install (no documentation!) and it 
seems benign.

> >Any other info/suggestions?
>
>I finally got my MacOS X CD in and I have a utility that can format and
>dump the Disk Partition tables.  Maybe this will give someone a clue
>about how to build a CD that is bootable on different systems so we can
>make the appropiate changes to mkhybrid.  Is anyone interested in a copy
>of the dump?

If it's not too long I'd just post it to the list.


Signature held pending an ISO 9000 compliant
signature design and approval process.
h.b.hotz@jpl.nasa.gov, or hbhotz@oxy.edu