[180] in NetBSD-Development
Re: Usage of NetBSD system packs
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Nov 9 07:04:08 1994
To: jhawk@MIT.EDU
Cc: netbsd-dev@MIT.EDU
In-Reply-To: Your message of "Wed, 09 Nov 1994 05:51:38 EST."
<199411091051.FAA03527@lola-granola.MIT.EDU.MIT.EDU>
Date: Wed, 09 Nov 1994 07:03:36 EST
From: Greg Hudson <ghudson@MIT.EDU>
> Now that 1.0 is out, to what extent do we want to put it on the new
> NetBSD system packs? That is, right now both the NetBSD and Linux
> system packs appear to have only /usr/athena; isn't the idea that
> they should be able to hold essentially the whole OS? (or is this an
> Athena paradigm that we're not following due to explicit policy
> decisison, or pragmatic lack of space?).
It's an Athena paradigm that I personally don't think we should
follow. It's a paradigm that (a) takes up a lot of disk space, (b)
tends to lead to a lot of buggy copies of setuid sendmail executables
in the sipb cell, (c) doesn't do anyone a whole lot of good, and (d)
leaves open the possibility that people will depend on the sipb cell
for basic system utilities, which would be poor.
> 1) Now that there's a stable release out, I'm willing to put
> together a small FAQ/procedure/inst disk for installing NetBSD 1.0
> from Athena over the 'net. Is someone else working on this?
No one's working on this that I know of, and it's somewhat crucial to
the success of the IAP class, so this would be a good thing.
> I do note that pts mem alternates telling me all the people in the
> group, and no one in the group, so perhaps Greg was bitten by that
> bug?
Uh, yeah (rosebud's protection database is screwed up), plus
/afs/sipb/project/netbsd's permissions.
I've modified /afs/.sipb/system/i386_nbsd1/usr/athena to be writable
by system:netbsd-dev and not system:gsipbbin.
