[Top][All Lists]

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

Re: bit-split, or: the schizophrenia of trusted computing

From: Marcus Brinkmann
Subject: Re: bit-split, or: the schizophrenia of trusted computing
Date: Mon, 01 May 2006 06:12:46 +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 Sun, 30 Apr 2006 23:47:22 -0400,
"Jonathan S. Shapiro" <address@hidden> wrote:
> On Mon, 2006-05-01 at 05:22 +0200, Marcus Brinkmann wrote:
> > > The first corresponds to one user making a "loan" of the space to the
> > > second. If I make you a "loan" of my right to storage, this does not
> > > imply that I should be able to read what you put in that storage.
> > 
> > Right, it does not.  However, it's not hard for you to encrypt the
> > data on write out.
> I think that you have "I" and "you" reversed. But also you are confused.
> This is a persistent system. There is no writing data out. The space
> that I have "borrowed" is part of my *heap*.

Jonathan, in the last days, everytime you think I am confused, I am
actually thinking clearly, and you just misunderstood.

I suppose you keep track of which memory frames you loaned and which
not.  You only write encrypted data to the loaned frames.  Problem

> Your ability to destroy the storage that I borrow from you does not
> create a moral hazard. I may lose my bits, but I knew that when I
> decided to use your storage.

If it creates a moral or otherwise hazard depends on the use case.

> This is completely orthogonal to whether I should be required to
> disclose my bits to you, or take expensive measures to protect them.

Yeah, so what?

> > I am mentioning this to draw attention to the fact that the real real
> > world allow for much wider range of nuances than a superficial
> > analysis suggests.  This should make one suspicious if the simple
> > technical means have the right properties, especially if they can be
> > enforced rigorously.
> That is all very nice, but I notice you did not address my point at all:
> the party making the loan (and I agree that it isn't quite a loan,
> because it can be instantly reclaimed) requires the right to reclaim,
> but not the right to read.

I explained to you why I do not think that this is necessarily the
case.  It is the case in your example only because you _said_ it is
the case.  But once your abstract descriptions hit the real world, you
don't know what happens.

> > > From his note, I believe that what Marcus is trying to disable is the
> > > ability to own *information*, which is different from storage.
> > 
> > Actually, both are separable concerns.  The struggle for freedom of
> > information is a struggle for free culture.  It is of secondary
> > concern here in this discussion.  The struggle for free hardware is a
> > conservative cry: I should be in control over how the bits flow in my
> > computer (yes, even if I choose to store information on it that is
> > owned by somebody else).
> Marcus: that is beautiful rhetoric. Does it actually *mean* anything?

It does mean something to me.

> > Well, first, there is a very obvious difference that can matter.  The
> > painting is not digital data.  Digital data has the peculiar property
> > that it can be copied and distributed to everybody who wants to have
> > it without loss of quality, quantity, and without marginal costs.
> I agree that this is true, but it does not seem relevant to the topic of
> conversation, which was "control over information".

I am surprised.  "control over information" is not my topic here.
Maybe you posted in the wrong thread?

Now I understood why I thought that the painting example was kinda
out of place.

I was talking about ownership of storage.

Anyway, I already answered the painting question.


reply via email to

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