From: Harvey Newstrom (mail@HarveyNewstrom.com)
Date: Fri May 05 2000 - 19:49:14 MDT
"Eugene Leitl" <eugene.leitl@lrz.uni-muenchen.de> wrote on Friday, May 05,
2000 8:57 PM,
> Not really, it is more like keeping mirrors of a server, to be able to
> smoothly cross over in case the main one fails. Multiple redundancy is
> an ancient tradition in mil/aerospace apps.
>
> I'm sorry to resort to hackneyed stereotypes, but IT people are
> familiar with computers, so it is probably a valid analogy. It is not
> the real thing when discussion uploaded copies, but it does come
> pretty close.
This is a good analogy which works for me. This sounds like a RAID array of
disks, all mirrored to keep the same data, with duplicate input and output
peripherals. I think I agree with you. In this case, I would claim that
you only have one computer or OS running. You have duplicated all the
software and hardware, but by keeping them in synch you only have one system
with redundant parts.
I think the semantics of the words are confusing us. I agree with your
descriptions of the examples, and your choices about what valid actions are
allowed. You call this killing a copy. I would call this removing
redundancy from a single entity. I think you would agree with my examples,
and my choices about what valid choices are allowed. Where I would consider
having a copy that could be killed, you would see diverged copies that are
no longer true (exact) copies. I think we are agreed on both examples, when
to allow copy deletion and when not to. I think we just disagree on the
labels of these examples, to determine which examples should be called
"killing a copy."
-- Harvey Newstrom <http://HarveyNewstrom.com> IBM Certified Senior Security Consultant, Legal Hacker, Engineer, Research Scientist, Author.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:28:26 MST