Re: Cohen's paper on byte order

From: Mok-Kong Shen (mok-kong.shen@t-online.de)
Date: 04/08/03 

From: Mok-Kong Shen <mok-kong.shen@t-online.de>
Date: Tue, 08 Apr 2003 10:18:32 +0200

Eugene Starokoltsev wrote:
>
> Bryan Olson <fakeaddress@nowhere.org> wrote in message news:<b7kka.278$IB6.12@newssvr16.news.prodigy.com>...
> > Eugene Starokoltsev wrote:
> >
> > > Well, suppose in some physical experiment I have got a table of 17325
> > > numbers, first of wich is 1.54086e-23 which I'd like to transmit to my
> > > colleagues secure. How do you suggest to improve the AES to make it
> > > useful from your point of view for this simple example? ;)
> >
> > Just as stated of course: define it on octet strings.
>
> Why I need to know detais of format my favorit programs? I use
> Microcal Origin and my colleague uses Matlab. If I transmit the file
> my e-mail attachment but my colleague can't open it - it is not a
> problem of my stupidity but the problem of e-mail protocol. If
> programmer adding AES to a byte-level protocol can't read the AES
> specification correctly to understande that in needs some glue - it is
> not a problem of programmer's stupidity but the problem of AES.
>
> BTW any octet stream can be replesented by bit stream.
[snip]

It isn't the problem of AES either! FIPS-197 gives
in Fig.2 the definition of the content of all 16 bytes
unambiguiously. If an implementor handles these bytes
in wrong ways in a transmision protocol (like reversing
the order of the bits) such that what a sender sends
arrives at the receiver differently, then that clearly
and certainly is in no case a fault of the AES
document in my view. (It looks that some bad programmers
who have done erroneous jobs would like to wipe away
their errors in accusing AES of defects where there
are actually none.)

M. K. Shen

Relevant Pages

  • Re: Rijndaels algorithm - clarifications
    ... Is Rijndael's algorithm involves only one key and actual data ... AES is a block cipher. ... you need a protocol which addresses the security features you ... uses a symmetric cipher for confidentiality (AES ...
    (sci.crypt)
  • Re: Cohens paper on byte order
    ... >> mapping be specified in AES itself or in the specification for the ... >> communication protocol you have mentioned just above, ... > but cannot tell you how to order the AES bits in each octet. ...
    (sci.crypt)
  • Re: Wireless security
    ... > Any links to the proposed stardard? ... >> Will these new cards support the new, AES based protocol? ... >> getting link layer encryption for a wireless LAN today, ...
    (sci.crypt)
  • Re: Security using AES
    ... >> design or implementation of such a protocol, so you need to know what ... > We would be elated to find an existing protocol. ... > and a random spoiler that makes the concatenation into an even AES block ... networks where an AES based mode known as CCM mode has been designed. ...
    (sci.crypt)