Subject: Re: Booting sd0 q(disk geometry versus bios geometry)
To: None <port-i386@NetBSD.ORG>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: port-i386
Date: 07/10/1998 10:57:51
>Which in fact means that the current 1.3 geometry system is crap.

Perhaps. The previous MBR geometry-handling code caused corrupted MBRs
for many, many people. If you didn't have a backup, you could well
lose data.  If it's still losing in 1.3.2, then there may still be
latent bugs in sysinst or in disklabel.

"Send a PR".


>Now with 1.3* I am struggling real hard with the system to get it booting
>from a 100MB disk that happens to have 1200 cylinders *sigh*

That sounds painful.  But have you tried doing that with the 1.2
install tools?  Do the 1.2 install tools do a better job of handling
an MBR geometry that cannot address the the NetBSD MBR-partition?

>While this compatibility thing with other OS etc. is nice if you 
>want to use more than one OS on disk, it is crap when using the entire
>disk for NetBSD.

But there is, apparently, hardware out there which won't boot properly
unless you have an "approved" geometry, and leave the first cylinder
free.  That's quite aside from the impact on someone who installs
NetbSD and subsequently decides to install a second OS on the disk.



>And - for user that are even less experienced than me - this is a reason
>to abort the NetBSD install and try another *BSD or Linux.

Perhaps so, but this was even *more* true with the old install method
and the "take a pencil and paper and have a calculator handy"
instructions.