Subject: Re: bin/466: i386 disklabel STILL clobbers DOS part. table!
To: None <firstname.lastname@example.org>
From: der Mouse <mouse@Collatz.McRCIM.McGill.EDU>
Date: 09/13/1994 07:13:25
>> I believe that this PR is incorrect.
>> The `c' partition is supposed to correspond with the NetBSD
>> partition in the DOS partition table.
>> The `d' partition is supposed to be the whole disk.
> I still have one of my drives labeled the same as it was under 0.8
> and probably 386bsd 0.1 before that and it does have the partitions
> "reversed" at least by your above definition... And I know that the
> drives were labeled like that for a reason. [...] I have a strong
> feeling that you're right, but am curious why the hell I've been
> doing it backwards for so damn long...
Perhaps because "traditionally" (ie, when not sharing disks and such),
c has been the whole disk, with all seven other partitions available
for "real" use? (Personally, I find this tendency to attribute
semantics to some partition letters quite annoying. It's bad enough to
lose 1/8 of the available partitions to a whole-pack partition, and I
recently ran into another system that hijacked paritition h for
something, and what's more, only raw partition h - block partition h
was what it "should" have been. I'd much rather provide different
devices, or bump things up to 16 partitions per disk....)