From owner-freebsd-current Sun Nov 24 21:03:00 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id VAA04698 for current-outgoing; Sun, 24 Nov 1996 21:03:00 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id VAA04686 for ; Sun, 24 Nov 1996 21:02:51 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id PAA17951; Mon, 25 Nov 1996 15:59:42 +1100 Date: Mon, 25 Nov 1996 15:59:42 +1100 From: Bruce Evans Message-Id: <199611250459.PAA17951@godzilla.zeta.org.au> To: bde@zeta.org.au, rgrimes@GndRsh.aac.dev.com Subject: Re: 2.2-ALPHA install failure Cc: current@freefall.freebsd.org, darrylo@sr.hp.com, jkh@time.cdrom.com Sender: owner-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> >Dedicating one drive, does not mean dedicating all drives, Microsoft's >> >install procedures are known for there ``I want the whole world'' phylosophy >> >and can cause people great pain if installed with on another disk in >> >the same systems as one of the FreeBSD ``bogus'' partitioned disks. >> >> Surely this is only caused by a installer error? I haven't used W95, >> but older versions of W can be installed in any directory on any hard >> disk and don't seem to touch other disks or directories (except for the >> usual things in the root directory). > >No, this is not a user error, this is the automagic of Windows 95 and >the design of the Microsoft installation tools. If it finds unclaimed >disk space per the MBR it will claim it, fdisk it, and format it for >you, all without asking you if it is okay to do this. This may have >been fixed in Win95/SR2. Bust surely it only installs on the drive(s) that you tell it to? >I have shipped 100's of systems with ``dedicated disks'' and every single >one of them had a valid MBR as far as start and size. Every OS I have >seen and worked with obey these two parameters, infact they favor them >over the start CHS, end CHS. I am not so concerned about bogus C/H/S >values in the MBR as I am about start/size values. Except W95. It doesn't honor the start and size parameters. A start of 0 places the MBR inside the slice. This should prevent all OS's except the one in the slice from modifying the MBR. A size of 1 should work just as well as a size of 50000. I don't know of any fdisks that actually honor a 0 start. This shows that a 0 start is invalid. Bruce