Bare-metal bootstrapping the Apple /// over the serial port
schmidtd
subconscious. Finally the solution appeared to me the way it often
does with my subconscious working as a detached process.
Anyway: it finally worked tonight. There's so little firmware in
the /// that I require you to type in a grub bootstrapper of 76 bytes
into the barely functional monitor, but once that's done, ADTPro does
the rest. It streams in a bigger loader, then a patched SOS kernel.
As part of the kernel's initialization process, it requests the
"interpreter" (ADTPro itself), then the SOS drivers over the serial
port rather than the disk. An upcoming update to ADTPro will also
allow the /// to format floppy disks (another one of those
subconscious things) so all /// computers rescued from the dustbin of
history will be able to rise again.
After re-seating all the chips, of course. ;-)
mdj
> After re-seating all the chips, of course. ;-)
At which point in time, the recommended way of doing this will eat
away at your subconscious ...
:-)
David Schmenk
Very clever. You must be the premier Apple /// coder on the planet now.
Which kind of brings up another question: How many Apple ///s are
still active? Whenever they show up one ePay, they seem to fetch a
pretty good price.
> After re-seating all the chips, of course. ;-)
Luckily I've never had to so this. The thought of applying the
recommended procedure (lifting the unit up 1.5 inches and dropping it)
makes me cringe. I could see the combined weight of the ///, Profile
and monitor crushing the table and send it crashing through to the floor
below. Opening the case and pressing the chips back in seems like a
safer plan.
Bill Garber
"David Schmenk" <dsch...@YUCH.gmail.com> wrote in message news:nQxxk.21081$mh5....@nlpi067.nbdc.sbc.com...
I was under the impression the prior was to reseat the IF cards. ;-)
Bill Garber from GS-Electronics
http://www.garberstreet.com
schmidtd
> schmidtd wrote:
> > Anyway: it finally worked tonight. There's so little firmware in
> > the /// that I require you to type in a grub bootstrapper of 76 bytes
> > into the barely functional monitor, but once that's done, ADTPro does
> > the rest.
maybe - in a field of one. :-) I've learned some neat tricks after
spending some quality time with the SOS kernel listing. I've never
been a particularly crafty assembly programmer - those folks worked
magic, as far as I'm concerned. Of course, they had some pretty cool
hardware backing them up.
> Which kind of brings up another question: How many Apple ///s are
> still active?
Six. Probably. :-)
> > After re-seating all the chips, of course. ;-)
>
> Luckily I've never had to so this. The thought of applying the
> recommended procedure (lifting the unit up 1.5 inches and dropping it)
> makes me cringe. I could see the combined weight of the ///, Profile
> and monitor crushing the table and send it crashing through to the floor
> below. Opening the case and pressing the chips back in seems like a
> safer plan.
I actually had some pretty serious trouble with the one I'm
borrowing. I removed the mobo and pulled, cleaned and reseated each
chip to get it working reliably when I was in the heat of the ADTPro
port a couple of months ago. During this recent kernel exercise, I
had to drop the pan and work the chips around a bit to get it under
control again, too. I'd never dream of using the drop (myth?)
procedure, myself. But it _is_ a pain to unscrew 12 screws to get at
the good stuff. I only have two holding the pan on right now. :-)
I can't put my finger on it now, but I saw a video of a forum recently
where the designer of the /// case was sort of defending himself. He
swears up and down it wasn't the thermal nature that was the cause of
the pain, but instead the closeness of the circuit board traces (which
was the last to be laid out by hand at Apple).
Michael J. Mahon
Nonsense. It was cramming too many watts into too little surface
area, with no possibility of air exchange with the outside.
The engineers were appalled that the case was designed before the
contents.
-michael
AppleCrate II: An Apple II "blade server"!
Home page: http://members.aol.com/MJMahon/
"The wastebasket is our most important design
tool--and it's seriously underused."
Steven Hirsch
>> Which kind of brings up another question: How many Apple ///s are
>> still active?
>
> Six. Probably. :-)
I'll see your six and raise you to eight: I have a /// and a ///+ functional
in my collection.
schmidtd
> schmidtdwrote:
> > I can't put my finger on it now, but I saw a video of a forum recently
> > where the designer of the /// case was sort of defending himself. He
> > swears up and down it wasn't the thermal nature that was the cause of
> > the pain, but instead the closeness of the circuit board traces (which
> > was the last to be laid out by hand at Apple).
>
> Nonsense. It was cramming too many watts into too little surface
> area, with no possibility of air exchange with the outside.
>
> The engineers were appalled that the case was designed before the
> contents.
>
> -michael
speaking at the Computer History Museum - Speaker Series 2007 (June 4,
2007). Here's the YouTube video:
http://www.youtube.com/watch?v=--7Br07QKMk
He starts speaking at 8:40, starts talking about the III at 23:15.
"It's not my fault": 25:45 and 26:30.
Michael J. Mahon
I'll take a look.
BTW, the problems resulting from the large thermal excursions may
well have been partly an issue with the PC board, but the actual
size of the excursions was a packaging issue.
The /// was an early example of Jobs' desire for absolute control
over the physical design leading to big problems for the engineers.