Path: utzoo!utgpu!water!watmath!clyde!rutgers!princeton!udel!rochester!bbn!uwmcsd1!ig!jade!ucbvax!sdcsvax!ucsdhub!hp-sdd!hplabs!hpda!hpihoah!hpirs!hpisoa2!spuhler
From: spuhler@hpisoa2.HP.COM (Tom Spuhler)
Newsgroups: comp.unix.wizards
Subject: Re: Jerry Pournelle on UNIX (From BYTE)
Message-ID: <1020002@hpisoa2.HP.COM>
Date: 11 Jan 88 17:23:06 GMT
References: <1495@osiris.UUCP>
Organization: Hewlett Packard, Cupertino
Lines: 21

>Yes, a write-protected root is a trivial one-line change to the kernel.  But
>that's not what we were considering; it still needs to have the functionality
>of a normal root -- that is, all the things in /etc could still be "updated,"
>for example; but how this could be done is not clear.  The Air Force kernel
>moved all modifiable files off the root file system, but it was a maintenance
I don't know what all would need to write to root while single user,
and what would happen if they (?) couldn't, but "what if" (you've seen
the commercial) there were two /etc directores.  One would be a minimal
subset on the read only FS and there would be a link from /etc to say,
/letc.  This copy would be used in single user mode and the like.
Another, full copy would be kept on another FS and would be mounted on
/etc  when the normal mounts occur  (going to init 2, for example). Now
all the stuff in /etc would be available and things would go on from
there.  The single user /etc would still be accessable via /letc. 
Something that comes to mind is how do you modify your /etc/checklist
(or BSD equiv.) when it's on CD ROM.

Actually, this would seem to have even more potential as a method to
unload a bunch of the stuff in /etc off root...hummmm...

Tom spuhler, Hewlett Packard Systems Performance Laboratory