On Thu, Feb 01, 2001 at 11:35:00AM -0800, James T. Nelson wrote:
> How exactly do cmd-opt-P-R without erasing teh System Disk patches?  How,
> exactly do you set up the machine?

Sorry, that was a typo. I mean cmd-opt-O-F.

> I tried resetting parameter ram about 4 times, then went into Mac OS 8.6,
> then ran System Disk with the patch described for macppc, set /chaos/control
> and kbd, then rebooted into OF with auto-boot? false.  Still no display,
> though i have blind kbd input.

Well, you *did* hack System Disk's resource fork as described in the
FAQ, right?

If not, do that. It didn't make my monitor sync every time, but I
get it the majority of the time now.

(Even that didn't help a couple of 7300s I have, though.)

All of these machines run headless for me, so I'm pretty unconcerned
except when I really *need* a console, and that happens extremely
seldom.

> I can run macppc fine over ttya, and can telnet to it from by i386.  Works
> great.  I have been running a numerical model through telnet.  NFS works
> fine, etc.

Well, you might consider just leaving it that way. There's no
reason a machine *needs* a console, unless you want a console to
use.

If you've got another computer to ssh (you really ought to ditch
telnet, btw) in from which you're used to using as your desktop
machine, stick with it.

The macppc port is only ready for desktop prime time on much more
recent versions of OF. Making these OF 1.0.5 machines good for
console use is really difficult, if not impossible. OF is just too
buggy and unreliable.

> >ofb0 at pci1 dev 11 function 0: Apple Computer Control
> >ofb0: 640 x 480, 8bpp
> >wsdisplay0 at ofb0: console (std, vt100 emulation), using wskbd0
> 
> 
> This is where mine differs.  See below.
> 
> >> "wsdisplay0 at ofb0"

You mean that?

As I said before, that doesn't make sense to me. The wsdisplay
driver shouldn't just not make any comments about creating a
console. But, maybe I should go inspect the code before shooting my
mouth off about it.

> Is there something in sysctl that needs setting?

Not that I can think of.

> Right.  It does differ.  That is what is interesting.  The boot process
> fails to set up the display at this point, I believe.
> The above is transcribed by hand.  I can copy the file directly from zterm
> log file, and ship it to you.  I will check again.

It seems like there should at *least* be a trailing colon after
"ofb0"... otherwise, wsdisplay shouldn't even have matched.

Again, I would need to look at the wsdisplay psuedo-device code to
really know what I'm talking about, and I never have.

> >(I've booted ornery OF 2.x 6500s this way, and I *think* I've done
> >it with OF 1.0.5 machines, but I'm not positive.)
> I agree, it should be possible.

Very curious.

I wish I could be of more help, but it's kind of hard to diagnose
this stuff remotely, even for someone who knows exactly what
"should" happen (which I don't, I just know what I've seen).

       ~ g r @ eclipsed.net