At 2:43 Uhr +0100 17.12.1998, Michael R. Zucca wrote:

>>I found it peculiar that the serial ports (2!) are probed (?) just fine but
>>cannot be used. On second thought, it may well be that the docks have their
>>own set of peripheral chips and the Duo's internal ports are disabled when
>>it is docked.
>
>The I/O chips are split between the Duo and the dock if I remember rightly.
>The serial and iwm probably really do live on the Duo itself. I think the
>only thing on the dock is video, NuBus, and maybe another SCSI controller.

>From looking at the Duo Dock II block diagram, it is the other way round: A
docked Duo provides internal video, SCSI for the internal disk, ADB, sound
and (optionally) the modem. All other I/O (floppy, serial, ethernet,
external SCSI/video, Nubus) is provided by the Dock.

>We should easily be able to support a docked machine. The devices all appear
>in Slot $E (See the Apple Tech specs for the dock).

Ahh... thanks for the pointer! I had the pdf files lying around, lost
somewhere on a 2 G disk.

So - that would mean I could probe "slot" $E and attach a "duodock" bus
(like "obio") to it which has devices attached whose drivers are available
anyway? I am just beginning to like the NetBSD autoconf framework...

>"Hot-swapping" a machine is a different matter. Since NetBSD does match/attach
>once at boot, I'm not sure if it's possible to do something like switch
>between
>the console at the LCD and the console at the Dock and vice-a-versa. Maybe a
>daemon could check a "dock device" and then load the appropriate kernel
>modules from user mode?
>
>Still, just getting a machine which is "steady state" docked probably wouldn't
>be that hard.

>From the Apple docs, you can only swap a sleeping Duo. I am not at all sure
we support putting a Powerbook to sleep (and waking it up again).

	hauke


--
"It's never straight up and down"     (DEVO)