l4-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Retracting the term ownership (was: Re: Separate trusted computing d


From: Marcus Brinkmann
Subject: Re: Retracting the term ownership (was: Re: Separate trusted computing designs)
Date: Sat, 02 Sep 2006 00:11:38 +0200
User-agent: Wanderlust/2.14.0 (Africa) SEMI/1.14.6 (Maruoka) FLIM/1.14.7 (Sanjō) APEL/10.6 Emacs/21.4 (i486-pc-linux-gnu) MULE/5.0 (SAKAKI)

At Fri, 01 Sep 2006 17:04:37 -0400,
"Jonathan S. Shapiro" <address@hidden> wrote:
> 
> On Fri, 2006-09-01 at 20:34 +0200, Marcus Brinkmann wrote:
> 
> > Well, it would be an interesting trick if you could change the nature
> > of information by mere postulation of a believe about it.
> 
> But this is exactly the question at hand: *is* this the nature of
> information, or is it a reflection of a correctable technological
> limitation?

From an information theory point of view, a "trusted computing system"
is nothing but a black box, which contains some information, which we
can't measure, an input line and an output line.  Through the input
line, we can add information to the box, and through the output line
we can receive information from the box.

Note that any information that I can not (at all) extract through the
output box is for all practical purposes not there.  It's as if it
doesn't exist.

The information that I can extract through the output line is
unrestricted by the "trusted computing system".  I hope that you agree
that this information has all the properties that Jefferson attributes
to it.

So, all the information I can extract is of the kind of information
that Jefferson describes.  All the information that I can not extract
does not exist for any practical purposes.  It can not have influenced
the output, as then it would be part of the output and thus
extractable.

A rigurous treatment will of course consider output probabilities
instead of output messages.

Does a "trusted computing system" with only analog output change
anything?  Not really.  The analog output can be quantized.  It may
contain noise, but if it does, then this simply reduces the
information that we can extract.  The information that is lost in the
noise is information that by definition we can not extract and thus it
does practically not exist.

I really can not see how "trusted computing" changes anything about
the status of extracted information.  Maybe I am misunderstanding what
you are trying to say.

Thanks,
Marcus





reply via email to

[Prev in Thread] Current Thread [Next in Thread]