Cryptography-Digest Digest #158, Volume #10 Thu, 2 Sep 99 05:13:03 EDT
Contents:
Re: Schneier/Publsied Algorithms (Forrest Johnson)
Re: n-ary Huffman Template Algorithm (Alex Vinokur)
SIGABA / ECM Mark 2 (JTong1995)
Re: Which of these books are better ? (Jaap-Henk Hoepman)
Schneier/Publsied Algorithms (Mixmaster)
Re: Schneier/Publsied Algorithms (Ruud de Rooij)
----------------------------------------------------------------------------
Subject: Re: Schneier/Publsied Algorithms
From: Forrest Johnson <[EMAIL PROTECTED]>
Date: Thu, 02 Sep 1999 04:42:30 GMT
In article <7qjcb8$26ui$[EMAIL PROTECTED]> SCOTT19U.ZIP_GUY,
[EMAIL PROTECTED] writes:
> I THINK THE QUESTION WAS TO TECHNICAL FOR MR "BS" HIMSELF
>THE GUY ASKED FOR TEST VECTORS. SURELY YOU HAVE SOME
>LIKE THE ONES I USE IN MY GLOAT CONTEST. THE KIND OF CONTEST
>THAT YOUR WEAK METHOD COULD NOT RUN USING ANY OF YOUR
>FAVORITE "FIPS" 3 LETTER CHAINNING METHODS.
>
> THE GUY EVEN CPITALIZED """"TEST VECTORS""" MORE THAN ONCE
>BUT I GUESS YOUR BRAIN OVER LOOKED IT. MAYBE THE USE OF
>CAPITAL LETTERS MADE IT TO HARD TO READ. HOW THE HELL DO
>YOU EXPECT HIM TO FIND THE ERRORS IF YOU CAN'T SEEM TO HAVE
>THE ABILITY TO ANWSER DIRECT QUESTION PUT TO YOU. DO YOU
>HAVE """TEST VECTORS""" FOR THE GUY OR NOT???
>
> SURELY THE "nsa" COULD SUPPLY SOME THAT WOULD SATISFY EVERY
>BODY OR ARE YOU TRYING TO HIDE CERTAIN FACTS. BECASUE I SUSPECT
>YOU CAN READ THE LETTER. AND THE LETTER PLAINLY ASKED FOR
>
>"""""""""" TEST VECTORS """"""""" HEY WAKE UP
>
Mr. Scott, I'm pleased to see that you are so passionate about people
answering questions asked of them. Perhaps you would be so kind as to
answer the questions I posed to you last week in several different posts.
In case you've forgotten, you made a claim that you had changed software
in fielded weapons systems. I asked you to identify which systems these
were.
Given your excoriation of Mr. Schneier, I'm sure you are now eager to
avoid the "cast the first stone" stigma.
I did ask quite a few questions, so I might have overwhelmed you. I'll
start with one or two easy ones this time and we can go from there:
1) Did you change software in a fielded weapons system, yes or no?
2) If yes, what weapons system was it?
TIA
------------------------------
From: Alex Vinokur <[EMAIL PROTECTED]>
Crossposted-To: sci.image.processing,sci.math,alt.comp.compression
Subject: Re: n-ary Huffman Template Algorithm
Date: Thu, 02 Sep 1999 05:11:07 GMT
In article <[EMAIL PROTECTED]>,
Mok-Kong Shen <[EMAIL PROTECTED]> wrote:
> Alex Vinokur wrote:
> >
>
> > > > Mok-Kong Shen <[EMAIL PROTECTED]> wrote:
> > > > > A question just for my understanding: How can frequencies be
> > > > > non-numerical at all? If you have a number of frequencies and
have
> > > > > only their ordering according to magnitude but not know their
> > > > > numerical values, how can you expect to obtain a coding that
is
> > > > > optimal?
> > >
> > > > > So from Huffman algorithm point of view
> > > > > it is not important what type of cost is.
> > > > > The algorithm is using only operator< and operator+.
> > >
> > > This only reflects the fact that a Huffman tree can correspond to
a
> > > wide range of frequency distributions, i.e. that a Huffman tree
does
> > > not uniquely defines the frequencies of the nodes. But a
non-numerical
> > > data type 'by definition' can't support the operator +
> >
> > In C++ you can define the operator+ (and other operators)
> > for any class (type).
> >
> > > (excepting that
> > > + is often employed to represent string concatenation, so A + B
> > becomes
> > > aabb, if A is the string 'aa' and B is the string 'bb'). Could you
> > show
> > > (with a concrete real-life example) an instance of a
'non-numerical
> > > cost', i.e. the 'value' of a variable of that type and the result
of
> > > applying your operator + on two such 'values' so that one may
better
> > > comprehend the semantics of that operator?
> >
> > See special page of DevCentral Learning Center:
> > http://devcentral.iftech.com/learning/tutorials/c-cpp/cpp/7.asp
>
> You have not answered my question. I was not asking how one can
> implement an operator if one has the proper semantics. I was asking
> about the semantics itself. In fact, the question involves matters
> at the most basic level. A data type by definition consists of
> a (finite/infinite) set of values and the operators that can
> work (operate) on these values. Now I know till now with my humble
> knowledge only values of the data type 'numerical cost', like 500
> dollars, 20 francs, etc. Can you give similar examples of values of
> the data type 'non-numerical cost'??? (I am fairly convinced that
> the majority of readers also haven't heard of 'non-numerical costs'
> before.) Without knowing the set of values underlying a data type,
> there is no sense to talk about defining an operator for that data
> type, let alone to write a program to process variables/constants
> of that data type.
>
> M. K. Shen
>
RMelter <[EMAIL PROTECTED]> wrote (sci.image.processing, 1999/02/22):
http://www.deja.com/=dnc/[ST_rn=ps]/getdoc.xp?AN=447218924
>
> Dear Alex:
>
> Some time ago I wrote some papers about
> Boolean valuations and metrics which might be thought of as a kind of
cost. Two
> of the references are:
>
> Boolean Valued Rings and Boolean Metric Spaces ,Arch.Math
15(1964),3534-363
>
> Boolean Distance for Graphs,Discrete Math. 39(1982),123-127 (with
> I.Tomescu,F.Harary,and U.Peled)
>
> Robert A. Melter
> Professor Emeritus
> Long Island University
Alex Vinokur wrote (1999/02/24):
> Walter Roberson wrote:
>
> > In article <7au9ab$7ot$[EMAIL PROTECTED]> you write:
> > Newsgroups: comp.theory,comp.compression,sci.math,sci.crypt
> > :So from Huffman algorithm point of view
> > : it is not important what type of cost is.
> > :The algorithm is using only operator< and operator+.
> >
> > I would think of it not as the '+' operator, but as the 'and'
operator,
> > where 'and' is used in the sense of grouping.
>
> Yes, it is intersting.
> Note. If we are using the C++ language we must use the
"operator+"
> term.
> But we can determine it (i.e. the operator+'s body)
> according to our intersts.
>
> >
> >
> > Using '+' tends to imply that arithmetic of some form can be used
> > on whatever the type is. Really though, all that the '+' has to do
> > is a set union. All the hard work is then left to '<' : is this set
of
> > objects ordered higher or lower than this other set of objects.
> >
> > Consider for example running the huffman algorithm over a set
> > consisting of a giraffe, a cow, an orange, and a tulip. Don't assign
> > weights to these, just assign a decision algorithm that can order
> > subsets of them. The first pass would determine whether giraffe <
cow
> > or cow > tulip and so on, and would pick the two lowest and would
take
> > their union. Suppose it was cow and tulip. You then consider
> > {cow, tulip} compared to {orange}, {giraffe} and {orange, giraffe}
> > and so on.
> >
> > This isn't technically different than just defining operator+ as
> > set intersection, but mentally it is different: it makes it clearer
> > that there doesn't have to *be* any way to "add" the values.
>
> Your idea and approach are very interesting.
> And even in this case we don't have to change/update the Huffman
> Template Algorithm.
> Only operator< and operator+ should be determined by
> user/designer/researcher.
>
> I think it is also worth sending to the newsgroups.
>
> Thank you very much.
>
> Alex Vinokur
>
>
Alex
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (JTong1995)
Subject: SIGABA / ECM Mark 2
Date: 02 Sep 1999 03:12:41 GMT
Does anyone know of an accurate computer simulation of the ECM Mark 2? I've
seen several Enigma implementations, but not a SIGABA.
Jeffrey Tong [EMAIL PROTECTED]<Jeffrey Tong>
PGP 5 Key available for download at WWW.PGP.COM Key ID: BFF6BFC1
Fingerprint: 6B29 1A18 A89A CB54 90B9 BEA3 E3F0 7FFE BFF6 BFC1
------------------------------
From: Jaap-Henk Hoepman <[EMAIL PROTECTED]>
Subject: Re: Which of these books are better ?
Date: 02 Sep 1999 08:28:45 +0200
On 1 Sep 99 01:49:24 GMT [EMAIL PROTECTED] () writes:
> JPeschel ([EMAIL PROTECTED]) wrote:
> : If it's available as an export to your country, you should
> : really get the Dobbs crypto CD. The CD contains all
> : of the books you mention and a few more. It costs around
> : 100 US dollars.
>
> I believe it may be so available now, since some of the ads in recent
> issues of Dr. Dobb's no longer mention that it is limited to the U.S. and
> Canada.
Indeed it is. I received mine last month over here in the Netherlands.
Jaap-Henk
--
Jaap-Henk Hoepman | Sure! We've eaten off the silver
Dept. of Computer Science | (when even food was against us)
University of Twente | - Nick Cave
Email: [EMAIL PROTECTED] === WWW: www.cs.utwente.nl/~hoepman
PGP ID: 0xFEA287FF Fingerprint: 7D4C 8486 A744 E8DF DA15 93D2 33DD 0F09
------------------------------
Date: Thu, 2 Sep 1999 00:14:31 -0700 (PDT)
From: Mixmaster <[EMAIL PROTECTED]>
Subject: Schneier/Publsied Algorithms
Hello Bruce
How is it posible that some of zour published algorithms...2fish have bugs in zour
source code?
There are only two possible explanations for this:
1. A legitimate mistake was made...but no correction was ever published for it...or
have you published the correction on your site..
2. A deliberate bug was placed in your source code by some unknown person...
Which leads me to this point:
How can we in the crypto community EVER Trust any Publsihed Source Code without
extensive testing and debugging... I wonder if you thought of this.
Are there any published TEST VECTORS for your algorithms...and possibly other
Algorithms...which treat the algorithm as a black box...etc...do you know of any such
TEST VECTORS..
But please Bruce...explain to us How is it that there are bugs in your own published
algorithms...I did see some messages about this topic few months back..and have you
made any corrections to them
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
------------------------------
From: Ruud de Rooij <*@spam.ruud.org>
Subject: Re: Schneier/Publsied Algorithms
Date: 2 Sep 1999 09:53:36 +0200
Reply-To: *@spam.ruud.org
Mixmaster <[EMAIL PROTECTED]> writes:
> Are there any published TEST VECTORS for your algorithms...and possibly other
>Algorithms...which treat the algorithm as a black box...etc...do you know of any such
>TEST VECTORS..
The twofish paper available from www.counterpane.com contains test
vectors, as required (I think) for AES submission.
- Ruud de Rooij.
--
ruud de rooij | *@spam.ruud.org | http://ruud.org | http://weer.moonblade.net
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and sci.crypt) via:
Internet: [EMAIL PROTECTED]
End of Cryptography-Digest Digest
******************************
