[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: testing with powermac/oldworld (probs in miboot, debootstrap, kernel)



On Thu, Jun 07, 2001 at 01:00:33PM -0500, Stephen R Marenka wrote:
> I think I remember yellowdoglinux working on a graphical interface to
> miboot, I wonder if anything came of it?

graphical interface != useful

it already has a graphical interface, its just a useless one that
doesn't let you do anything. 

yellowdog doesn't use miboot either, they just require MacOS on
oldworlds.  

> I believe debootstrap 0.1.10 has a workaround fix which essentially
> includes a --force-confold to the above line. This is indeed a
> powerpc-utils--related problem. powerpc-utils is prompting about a new
> conffile but stdin is /dev/null, so it fails to configure.

powerpc-utils should still be fixed, that diversion needs to go away.  

> It would be nice to know what nvsetenv calls were made, this should be
> in /var/log/messages (via INFOMSG). It might also be useful to have the
> output of nvsetenv on your machine.

the way nvsetenv is being called now is broken, if ofpath fails
boot-device will be set null and the machine will fail to boot.  but
since he booted macos it means nvsetenv was never run at all.  

either that or it set boot-device null and OF considers that invalid
and reset itself to defaults.  

what we should be doing is this:

run ofpath, read its stdout into a variable, send its stderr to
syslog, check its return code, if its != 0 bail.  then strip off the
trailing newline from the ofdev variable and strcat a 0 onto it.  then
later we call nvsetenv boot-device %s, ofdev

-- 
Ethan Benson
http://www.alaska.net/~erbenson/

Attachment: pgpVjL4NcqEV_.pgp
Description: PGP signature


Reply to: