Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!uakari.primate.wisc.edu!sdd.hp.com!elroy.jpl.nasa.gov!decwrl!sgi!shinobu!odin!thestepchild!rhartman
From: rhartman@thestepchild.sgi.com (Robert Hartman)
Newsgroups: comp.unix.questions
Subject: Re: Norton Go Home (Revisited)
Message-ID: <1991Apr17.040303.16097@odin.corp.sgi.com>
Date: 17 Apr 91 04:03:03 GMT
References: <1081@keele.keele.ac.uk> <1991Apr16.182050.2028@odin.corp.sgi.com> <26480:Apr1622:26:2091@kramden.acf.nyu.edu>
Sender: news@odin.corp.sgi.com (Net News)
Organization: Silicon Graphics, Inc., Mountain View, CA
Lines: 20

In article <26480:Apr1622:26:2091@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
>In article <1991Apr16.182050.2028@odin.corp.sgi.com> rhartman@thestepchild.sgi.com (Robert Hartman) writes:
>> My main point had to do with the recognition that any persistent
>> condition that allows users to lose important data is either an
>> implementation bug or a flaw in the design.
>
>Probably because nobody knows how to make the system distinguish between
>important data and unimportant data. The only hardware that can satisfy
>your requirements---i.e., that does not let users lose data---is a WORM.
>
>---Dan

No Dan, you let the user tell you which files shouldn't be clobbered by
treating deletion/truncation as another type of file access and giving
it a separate permission mode--like I suggested.  Other OSs do this
(Tandem did this ten years ago), so it's not like I'm asking for the moon.

Well, I've said my say on this.  Perhaps I've said too much.

-r