From: Karl Lui Barrus <klbarrus@owlnet.rice.edu>
Date: Sat, 14 May 94 11:00:07 PDT
To: sinclai@ecf.toronto.edu (SINCLAIR DOUGLAS N)
Subject: Re: Message Havens
In-Reply-To: <94May14.025626edt.193@cannon.ecf.toronto.edu>
Message-ID: <9405141759.AA29918@flammulated.owlnet.rice.edu>
MIME-Version: 1.0
Content-Type: text/plain


SINCLAIR  DOUGLAS N wrote:
>Agreed.  I guess what I was trying to say (though I didn't actually
>come out and say it) is that the load is O(N^2).  While not a problem
>now, it would be nice to find an O(N) solution.  Or, more likely, 
>an O(Nlog(N)).

Okay, I think that you are overlooking something which appears obvious
(to me) but evidently it is not.  I am guilty of skipping over some
details.

The fact that you take the messages from a haven or several havens or
all the havens doesn't allow anybody to figure out if you are actually
reading any of the messages!

If message havens become popular, then (as you mentioned in a previous
post) there may be hundreds set up for people to use.  This would
decrease the load on each one.  If there many message havens to choose
from each would only have to serve a few hundred or so people (much
like banks).

Again, only if you want to hide your pseudonym, you can simply
download all the files and foil any logging on the part of the haven.

It is NOT necessary to download all the files from ALL the havens.
You just use your preferred one.  If you get all the files from the
haven you use (and ONLY the haven you use), you still leave the haven:

a) unable to determine your pseudonym
b) unable to figure out if any of the messages are indeed for you

The haven would not be able to figure out if anybody is communicating
with you since the haven can't be sure you are even reading the
messages you grab.

And if you do regularly receive messages at one haven, you can pick
another one, and get all the message from both.  And neither haven
would be able to determine what messages you are reading, if any at
all.  You could get even fancier and communicate back to your friend
via a different haven altogether, which you never go to.

So before doing a detailed complexity analysis, I think it is best to
think about the protocol a bit more.

-- 
Karl L. Barrus: klbarrus@owlnet.rice.edu         
keyID: 5AD633 hash: D1 59 9D 48 72 E9 19 D5  3D F3 93 7E 81 B5 CC 32 

"One man's mnemonic is another man's cryptography" 
  - my compilers prof discussing file naming in public directories