Duncan McEwan writes:
> So I started looking at what was necessary to change NetBSD's partition ID
> from 165 (0xa5).  I changed the code in sys/arch/i386/stand/lib/biosdisk.c to
> check first for 167 then 165.

167 is, sadly, used by NeXTStep.

I suggest looking at my new fdisk.c table to find an empty one.

I also suggest that, if you are going to go throught the trouble of
all of this, that you hack all the code so it takes EITHER MBR type --
that is, either the old or the new -- or takes just the new, depending
on a #define. We could then use this code to slowly transition to a
new MBR type. I dunno that everyone likes the idea, but it seems
sensible in the long run.

Oh, and on looking for uses of the old type, I suggest looking at
machine/disklabel.h to get an idea of the type number #defines that it
might be lurking under.

Perry