Subject: Re: Well, you may be... was Re: I am happy to report...
To: Bill Studenmund <skippy@macro.Stanford.EDU>
From: David A. Gatwood <marsmail@globegate.utm.edu>
List: port-macppc
Date: 11/11/1998 15:33:20
On Wed, 11 Nov 1998, Bill Studenmund wrote:

> Well, that's more a choice for Tsubai as he's working on the adb stuff and
> I'm not. :-)

The you was a collective NetBSD-ppc thing, really....  :-)


> It would be borrowing back. :-) Though since MkLinux boots after MacOS
> boots, I suspect they wouldn't see this problem as enough of MacOS would
> have run to readdress one of the keyboards to some other #. It's only when
> booting through Open Firmware that I think we'd run into this problem, so
> only we, linuxpmac, and OpenBSD would hit it.

MkLinux sees it, too, or at least it contains code to handle it (whether
it's necessary or not and whether it ever gets used or not is another
question).  Perhaps it resets everything back to its power-on state in one
way or another... I'm not sure, but there's definitely code in there to
look for multiple devices at the same address, attempt reassignment, and
an unresolved flag that it sets if it's not able to reassign it.  That's
all in the MkLinux adb.c file....


> > BTW, are there really two keyboards or is one of them your mouse?  Or is
> > there yet a third device lurking at address 2?  Just wondering....
> 
> It's part of the mouse. It's how it can do meta-key assignment under
> MacOS.

That's sort of what I figured.  We're having fun with that under MkLinux
at the moment, with lots of new problems caused by the addition of gpm
(mouse on the console) support....  The linux server has a hack to treat
it as a mouse until it sends a keystroke other than an arrow key -- which
works, but....


> As an aside, the readdressing needed here is why this mouse makes the
> debugger key stop working on my IIsi. :-) The IIsi adb code only
> recognizes the reset and debugger soft keys from adb address 2, not any
> valid kbd. The main keyboard (with the power button) usually gets moved to
> 15, so things break. :-(

That's interesting.  If there's a way to reset things to the power-on
state, maybe port-mac68k should swipe the MkLinux code, too, as the mach
kernel code appears to assign the real keyboard to ID 2 (at least no one
has found a fake keyboard there that I've heard about...).


Later,
David

David A. Gatwood                         Visit globegate's internet
dgatwood@globegate.utm.edu                  talker, Deep Space 36
http://globegate.utm.edu                telnet globegate.utm.edu:9624

-----BEGIN GEEK CODE BLOCK-----
Version 3.1
GCS/CC/FA/H/L/MC/M/MU/PA/TW d-@ s:>- a-- C++ ++>$ UBLAS*++ ++>$
P+?>$ L++ +>$ !E--- W++ +>$ N++(++ +)>++ +$ !o? K-? !w--- !O
M++>$ !V-- PS+>$ !PE- Y+>$ PGP+>$ t++ +>$ 5+>++ ++$ !X- !R tv+>$
b++>$ !DI !D- G++(++ +)>$ e>++ ++ h--! r--- !y-
------END GEEK CODE BLOCK------