Re: [dtn-security] Re: [dtn-dev] Re: SDNV-new

Rajesh Krishnan <krash@bbn.com> Thu, 26 May 2005 21:26 UTC

Received: from a.bbn.com (a.bbn.com [128.89.80.80]) by webbie.berkeley.intel-research.net (8.11.6/8.11.6) with ESMTP id j4QLQMV09725; Thu, 26 May 2005 14:26:23 -0700
Received: (from krash@localhost) by a.bbn.com (8.11.0/8.11.0) id j4QLQHq01249; Thu, 26 May 2005 17:26:17 -0400
From: Rajesh Krishnan <krash@bbn.com>
Message-Id: <200505262126.j4QLQHq01249@a.bbn.com>
Subject: Re: [dtn-security] Re: [dtn-dev] Re: SDNV-new
To: dtn-security@mailman.dtnrg.org
Date: Thu, 26 May 2005 17:26:17 -0400 (EDT)
Cc: dtn-dev@mailman.dtnrg.org
In-Reply-To: <20050526184825.GE4301@pisco.cs.berkeley.edu> from "Michael Demmer" at May 26, 2005 11:48:25 AM
X-Mailer: ELM [version 2.5 PL3]
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: dtn-security-admin@mailman.dtnrg.org
Errors-To: dtn-security-admin@mailman.dtnrg.org
X-BeenThere: dtn-security@mailman.dtnrg.org
X-Mailman-Version: 2.0.13
Precedence: bulk
Reply-To: dtn-security@mailman.dtnrg.org
X-Reply-To: krash@bbn.com
List-Unsubscribe: <http://mailman.dtnrg.org/mailman/listinfo/dtn-security>, <mailto:dtn-security-request@mailman.dtnrg.org?subject=unsubscribe>
List-Id: DTN Security Discussion <dtn-security.mailman.dtnrg.org>
List-Post: <mailto:dtn-security@mailman.dtnrg.org>
List-Help: <mailto:dtn-security-request@mailman.dtnrg.org?subject=help>
List-Subscribe: <http://mailman.dtnrg.org/mailman/listinfo/dtn-security>, <mailto:dtn-security-request@mailman.dtnrg.org?subject=subscribe>
List-Archive: <http://mailman.dtnrg.org/pipermail/dtn-security/>

> > I don't much care one way or another.  Do we really think we're
> > likely to need to represent numbers bigger than (2*68) - 1 in SDNVs?
> 
> Very doubtful if all they're used for is lengths, maybe so if they're
> used for other things like crypto keys and such. 
> 
> > If so, what convinces us that we're likely to need 16 bytes but not
> > equally likely to need 128 bytes?  Certainly we can use an encoding
> > scheme like this, or lots of others (for example, 1 2 4 8 16 32 64
> > 128), but the fact that we can doesn't necessarily mean we should.
> > What's the rationale for this particular system?
> 
> Well -- mostly a gut feeling at the time that we may find 12 and 16
> more useful than 5 and 7, but I confess that I don't have any
> particular proposed use cases or concrete justification for this
> feeling. So I'm really fine either way as well.

I do not know what possible uses are being considered for SDNVs in DTNs,
but it may be worth taking a look at some common sizes of interest.

              1B Bluetooth. (B = Byte).
              4B IPv4.
              6B Ethernet.
             12D for UPC (D = decimal digit).
             13D for ISBN.  
             13D for European Article Numbering.
             15D for E.164.
             16B IPv6.
             20B ATM End System Address.     > 16 octets.
            128B 1024-bit Cryptographic keys  
            512B 4096-bit Cryptographic keys and signatures 
                 are supported by several implementations

Best Regards,
Rajesh