Cryptography-Digest Digest #784, Volume #8 Tue, 22 Dec 98 00:13:03 EST
Contents:
Re: Computers getting faster? (Ralf Muschall)
Re: Forgotten Blowfish password (Mark)
Make Fast Random Number Generator? (Jim Trek)
Re: What is Randomness? (Lincoln Yeoh)
Re: Computers getting faster? ("Trevor Jackson, III")
Re: What is Randomness? (Darren New)
Re: coNP=NP Made Easier? (rosi)
Re: DIRT ? ("donoli")
Protocol needed for exam grading.... (Patrick Juola)
Re: Protocol needed for exam grading.... (Jonah Thomas)
Re: Protocol needed for exam grading.... ("Steve Sampson")
Re: Good s-boxes -characteristics thereof (Gramps)
Re: Make Fast Random Number Generator?
Re: RC4 in 8-bit vs 16-bit (Bruce Schneier)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Ralf Muschall)
Subject: Re: Computers getting faster?
Date: Mon, 21 Dec 1998 23:07:34 +0100
John Savard wrote:
> I always thought that the hardest part of making an optical chip was
> developing a switch by which one beam of light can influence another,
> and the article doesn't explicitly mention much of a breakthrough in
> that area. (Unless they're using something based on dye lasers, where
Unless some *very* great breaktrough happened in the last
few years, this is still the problem. The only process which
is fast enough to handle reasonable speeds is third-order
nonlinearity in transparent media (other processes involve
stuff like creation of electron-hole pairs, leading to time
delays and energy dissipation). Unfortunately the third-order
coefficients are rather small.
In addition, an optical waveguide needs a transversal dimension
in the micron-range (i.e. several times as large as a
transistor), and to get just *linear* coupling between adjacent
waveguides, one needs them in a length of the millimeter range.
I.e. even the simplest *passive* element is several thousand times
as large (in volume) as a transistor.
Ralf
------------------------------
From: [EMAIL PROTECTED] (Mark)
Subject: Re: Forgotten Blowfish password
Date: Mon, 21 Dec 1998 22:16:14 GMT
On 10 Dec 1998 19:05:08 -0500, [EMAIL PROTECTED] (Frank Gifford)
wrote:
>Perhaps there is already code out there. If not, it would not be
>difficult to write up the code for that since Blowfish is freely available.
>It may take a couple of hours to run, depending on the words and whether
>there are special things like capitalization: garbagecan is different
>than: GarbageCAN
>
>-Giff
My 'puter had an accident; i'm back now though.
I'm certain there are no upper case letters in the password.
Would you be good enough to write the code for me? Don't be
embarrassed to say "no." I have no idea if what i'm asking will take
five minutes or five hours so I certainly won't be offended.
Otherwise, where would the best place be to find such code (if it
exists)? IRC?
Thanks for the help,
Mark.
------------------------------
From: Jim Trek <[EMAIL PROTECTED]>
Subject: Make Fast Random Number Generator?
Date: Mon, 21 Dec 1998 17:23:56 -0500
Does anybody here make a fast random number generator or have the
capability to design and build one that will provide 2 million or
more bits per second for a PCI slot or a universal serial bus?
Jim Trek
[EMAIL PROTECTED]
http://www.frontiernet.net/~progress/
Future Beacon Technology
128 Main Street
Brockport, NY 14420
(716) 637-0256
------------------------------
From: [EMAIL PROTECTED] (Lincoln Yeoh)
Subject: Re: What is Randomness?
Date: Mon, 21 Dec 1998 20:01:44 GMT
Reply-To: [EMAIL PROTECTED]
On Fri, 18 Dec 1998 22:17:35 GMT, Darren New <[EMAIL PROTECTED]> wrote:
>My understanding (from outside the crypto area, tho) is that a string is
>random if (essentially) you cannot write a computer program shorter than
>the string that outputs that string. What it comes down to is that the
>only way to print the string is to embed it in the program.
But ten zeroes could be a pretty valid random number. It's just a rather
compressible sequence.
Link.
****************************
Reply to: @Spam to
lyeoh at @[EMAIL PROTECTED]
pop.jaring.my @
*******************************
------------------------------
Date: Mon, 21 Dec 1998 17:34:00 -0500
From: "Trevor Jackson, III" <[EMAIL PROTECTED]>
Subject: Re: Computers getting faster?
Ralf Muschall wrote:
> John Savard wrote:
>
> > I always thought that the hardest part of making an optical chip was
> > developing a switch by which one beam of light can influence another,
> > and the article doesn't explicitly mention much of a breakthrough in
> > that area. (Unless they're using something based on dye lasers, where
>
> Unless some *very* great breaktrough happened in the last
> few years, this is still the problem. The only process which
> is fast enough to handle reasonable speeds is third-order
> nonlinearity in transparent media (other processes involve
> stuff like creation of electron-hole pairs, leading to time
> delays and energy dissipation). Unfortunately the third-order
> coefficients are rather small.
> In addition, an optical waveguide needs a transversal dimension
> in the micron-range (i.e. several times as large as a
> transistor), and to get just *linear* coupling between adjacent
> waveguides, one needs them in a length of the millimeter range.
> I.e. even the simplest *passive* element is several thousand times
> as large (in volume) as a transistor.
>
> Ralf
Actually there has been exactly such a breakthrough. Quantum optics has
just be demonstrated. The experiment in question was based on multiplexing
32 channels through a quantum optical line that went past an optical
quantum point. Photons tunnel out of the line to the point.
This demonstrates the feasibility of optical layouts around 0.006 micron
IIRC (which I may not be). Apparently this recent demonstration is getting
a LOT of attention because it provides for pure optical rather than hyprid
electro/optical switching devices.
------------------------------
From: Darren New <[EMAIL PROTECTED]>
Subject: Re: What is Randomness?
Date: Mon, 21 Dec 1998 22:38:09 GMT
> But ten zeroes could be a pretty valid random number. It's just a rather
> compressible sequence.
Perhaps I should have phrased that differently. Replace the original
with
"A string is defined as random if...." in certain areas of computer
science.
Obviously, any sequence *could* be random, and that's what makes OTPs
work for cryptography. I'm just saying that asking whether some
particular string is "random" makes little sense without a description
of what "random" it has to fulfill. I wouldn't want to send a ten-byte
message with an OTP of "0 0 0 0 0 0 0 0 0 0" even if that's what came
out of my reverse-biased transistor. ;-)
--
Darren New / Senior Software Architect / MessageMedia, Inc.
The Three Scrooges: "Repent, Jacob Marley!" >poink!< Ow! "Woo-woo-woo!"
------------------------------
From: rosi <[EMAIL PROTECTED]>
Crossposted-To: sci.math,comp.theory
Subject: Re: coNP=NP Made Easier?
Date: Mon, 21 Dec 1998 12:47:50 -0800
Bryan Olson wrote:
>
> rosi wrote:
> [snip]
Dear Bryan,
Thanks for your reply post.
I think I am not confused and you are NOT ilias.
I can, at this time, only answer one person, focusing on one set of
questions and getting one thread taken care of. So if you would please
wait till the issue is settled with ilias, we could continue with your
questions and opinions.
One other thing, yet very important. Your interpretation and under-
standing of ilias's notions, opinions, statements, etc. are likely
very correct, and it will be appreciated when you try to help people
to explain whatever they mean. However, I need his words. If you had
joined the discussion earlier and given your opinion on 21, I likely
could have found an ally.
Would you commit to carrying out this through to the end once
we start? I.e. either we agree my argument is correct, or my
argument is faulty, or one side is shown inconsistent or contradictory
(if some simple questions are answered squarely and directly without
evasion)?
By the way, ilias perhaps has already seen that whichever notion
he uses, the issue IS settled (if ND is a well defined concept). He may
even have sensed something about the bigger picture. I believe he will
post again with his asnwers to those questions you expressed your
opinion on. I do not think that he is afraid to respond any more. So
please be patient. In the meanwhile, you can prepare your notion of
a NDTM for solving SS and post for our discussion. You may, of course,
choose one from 26 and 27, or give a precise one of your own (well-
defined assumedly). You may also get ready to answer the questions I
posed to ilias. For simplicity, you may answer in the following way:
1. YES
2. YES
3. NO
21. YES
etc.
Thanks again for your reply.
--- (My Signature)
P.S.
I believe your e-mail address is a real one. Mine is.
I think I have secured a second plumbing job for weekends and
would be extremely busy once started. If you do not mind, when you post,
would you cc onto [EMAIL PROTECTED]? I would really appreciate it. I would
also try to check sci.crypt as frequently as I can. --- (My Signature)
------------------------------
From: "donoli" <[EMAIL PROTECTED]>
Crossposted-To: alt.2600,alt.2600.hackerz,alt.hacker.learning
Subject: Re: DIRT ?
Date: Mon, 21 Dec 1998 12:59:18 -0000
NUTSA wrote in message <[EMAIL PROTECTED]>...
>On Sat, 19 Dec 1998 01:57:24 -0000, "donoli" <[EMAIL PROTECTED]>
>wrote:
>
>>
>>[EMAIL PROTECTED] wrote in message
>><75e00q$4lh$[EMAIL PROTECTED]>...
>>>Anybody know how to tell if you have the "DIRT"
>>>trojan installed on your PC. Also, what is the
>>>best way to remove it?
>>>
>>Try the Dirt Devil. It worked for me. Donoli.
>>
>>
>
>
>Now are you guys being facetious in your followups to this inquiry
>about the D.I.R.T. program??? Is there a program called Dirt Devil
>that will detect and remove this electronic surveillance crap??? This
>has caused me some concern since I surfed over the home site of the
>software company offering this to law enforcement agencies and the
>military following a lead provided by a ZD net article about online
>spy tools. The indications I got from the site was that something WAS
>being installed on my box. For those who do not know D.I.R.T. is
>similar to B.O. but a whole lot meaner and newer... Please Let me
>know if you will if there IS a detecter and remover...
>
>
I was only kidding about dirt devil. The site you're talking about is
thecodex.com, right? It's only a joke about something happening to your HD.
I thought the same thing untill someone pointed out the source code to the
web page. However, if you think there is a trojan on your HD, don't wait
for someone to write a detection program. Learn how to search the registry
to see if there's something that doesn't belong. Here's an example on how
to do that for BO http://www.iss.net/xforce/alerts/advise5.html Do the same
thing for DIRT. BTW DIRT was around before BO AFAIK. Donoli.
------------------------------
From: [EMAIL PROTECTED] (Patrick Juola)
Subject: Protocol needed for exam grading....
Date: 18 Dec 1998 10:40:20 -0500
I've got an interesting problem here at Duquesne. One of my students
just asked why he can't take the final over the Web. Ignoring the
obvious rejoinder "Because I didn't put the final over the Web,"
it raises the question of why I don't, or can't, put the final on
the Web.
A basic problem, of course, is security. How can I distribute
a document in such a way that I know that a) only the intended
recipient will get it, and b) only the intended recipient will be
the one who responds to the document. I'd like to prevent him
from sitting in a room with all his friends and collectively
brainstorming the answers to the questions. And possibly c) as well --
he only has the document for a fixed time.
The best solution I've been able to come up with unfortunately
requires hardware support.
-kitten
(p.s. And any response containing the phrase 'scott' in all caps
will be ignored. -k.)
------------------------------
From: Jonah Thomas <[EMAIL PROTECTED]>
Subject: Re: Protocol needed for exam grading....
Date: Tue, 22 Dec 1998 02:33:24 GMT
[EMAIL PROTECTED] (Patrick Juola) wrote:
>I've got an interesting problem here at Duquesne. One of my students
>just asked why he can't take the final over the Web. Ignoring the
>obvious rejoinder "Because I didn't put the final over the Web,"
>it raises the question of why I don't, or can't, put the final on
>the Web.
>A basic problem, of course, is security. How can I distribute
>a document in such a way that I know that a) only the intended
>recipient will get it, and b) only the intended recipient will be
>the one who responds to the document. I'd like to prevent him
>from sitting in a room with all his friends and collectively
>brainstorming the answers to the questions. And possibly c) as well --
>he only has the document for a fixed time.
This problem has self-defeating criteria. You might as well ask,
"Why can't I have a bathtub on the bus so I can take my morning bath
while I commute?" or "Why can't I just give blank checks to all my
suppliers and let them pay themselves so I don't have to think about
it?"
If it works to put your students entirely on the honor system, then
you can afford to put your tests on the internet. Otherwise the
internet is exactly the wrong place for them.
The way you keep students from getting help from all their friends
is to not allow communication during the test. The point of putting
tests in the internet is to allow them to look at the tests anywhere.
It just doesn't match up.
If you want to put them on the honor system, you could do some
more-or-less controlled tests to see how well it works. You could
announce that you may or may not have an additional in-class test
after the internet test, and that the internet test will be on the
honor system but the in-class test will not. Then if you actually
do an in-class test you could compare the results and see whether
there's a statistically significant difference.
If your students are honorable then you don't need any particular
cryptography to keep them from getting access to the tests at the
wrong time etc. And it won't matter who else looks at them, the
honorable students will refuse to listen to inside information.
------------------------------
From: "Steve Sampson" <[EMAIL PROTECTED]>
Subject: Re: Protocol needed for exam grading....
Date: Mon, 21 Dec 1998 20:42:24 -0600
You could give him the final over the Web, but to do it outside
of the exam room is completely wrong.
1) Without a proctor, the test is worthless. No one witnessed
the exam. Think about accepting credit from another school
who did tests via Web.
2) If the student flunks, he may claim that someone else has
modified his answers, or a bug in the software. Suppose he
sends in 5 different tests. Which one is the right one? He
says he must have hit the wrong keys. It's all your fault he
flunked the final.
3) The reason you don't give tests over the Web, is that you
don't trust the students to not cheat. There is no way you
can prevent cheating outside a testing room, or your view.
4) He can print it out at any time. But hell, he probably already
has a copy from 3 years ago when you gave the same test :-)
College is all about passing tests, and has very little to do
with the real world. Let him do the Web stuff after his degree.
Patrick Juola wrote in message <75dt14$h38$[EMAIL PROTECTED]>...
>
>I've got an interesting problem here at Duquesne. One of my students
>just asked why he can't take the final over the Web. Ignoring the
>obvious rejoinder "Because I didn't put the final over the Web,"
>it raises the question of why I don't, or can't, put the final on
>the Web.
>
>A basic problem, of course, is security. How can I distribute
>a document in such a way that I know that a) only the intended
>recipient will get it, and b) only the intended recipient will be
>the one who responds to the document. I'd like to prevent him
>from sitting in a room with all his friends and collectively
>brainstorming the answers to the questions. And possibly c) as well --
>he only has the document for a fixed time.
>
>The best solution I've been able to come up with unfortunately
>requires hardware support.
>
> -kitten
>
>(p.s. And any response containing the phrase 'scott' in all caps
>will be ignored. -k.)
------------------------------
From: Gramps <[EMAIL PROTECTED]>
Subject: Re: Good s-boxes -characteristics thereof
Date: Mon, 21 Dec 1998 18:54:20 -1000
Medical Electronics Lab wrote:
>
> Anthony Lineham wrote:
> >
> > What characteristics are generally thought to be
> > good for an s-box?
> > I know that the answer to this is algorithm
> > dependent, so how about wrt DES.
> > I understand the DES s-boxes are "optimised
> > against differential cryptanalysis."
> > What are the characteristics that make this the
> > case?
> You can read about it on
http://www.io.com/~ritter/RES/DIFFANA.HTM
In brief, the output bits of the S-Box can be XORed to make a
differential code. Some Output Differentials occurs more often than other
Oytput Diffurentulz,zo fur lukingut muni untup patenx wikd onecun often
keep thei if the iiaput numerz can add up. The DES s-bux iz nut uptymixed
for it toe!
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Make Fast Random Number Generator?
Date: 22 Dec 98 03:48:43 GMT
Jim Trek ([EMAIL PROTECTED]) wrote:
: Does anybody here make a fast random number generator or have the
: capability to design and build one that will provide 2 million or
: more bits per second for a PCI slot or a universal serial bus?
If you mean a generator of _true_ random numbers from a physical noise
source, the cheap conventional technologies are quite limited in speed. I
think you'ld have to go to a radioisotope source to do something like
this, but I'm *far* from being an expert in this area.
John Savard
------------------------------
From: [EMAIL PROTECTED] (Bruce Schneier)
Subject: Re: RC4 in 8-bit vs 16-bit
Date: Tue, 22 Dec 1998 04:27:37 GMT
On 21 Dec 1998 22:37:24 +0100, Anonymous <[EMAIL PROTECTED]> wrote:
>I just implemented CipherSaber-1 with the 8-bit RC4 state machine as
>described in their FAQ and everything is working properly. I'm on my way
>to becoming a CipherKnight... :)
>
>I started wondering what would happen to the relative strength of the
>cipher if the state table was instead a 65536-entry table of 16-bit words.
>The initial mix step would make MANY more passes through a given key (for
>example, a 54-char key with the 10-char IV would cycle 1024 times, not
>just 4 times), and I would think that a 16-bit state table would take far
>longer before it began repeating.
>
>In other words, would the 16-bit cipher be 256 times more difficult to
>break or would some other difficulty factor overrule that (such as still
>having the same keysize, say 80 bits of entropy)?
Thinking about RC4 variants with alternate alphabet sizes is certainly
an obvious question. I'm sure Rivest did some analysis, but there
isn't anything public.
My feeling about a 16-bit alphabet are mostly negative.
1. The weak point in RC4 seems to be the key schedule. It
It would take a LOT of key schedule to mix a 65536-entry
table. Such a key schedule would make the cipher very
inefficient for anything but large blocks of text.
2. Since the number of permutations of the 8-bit table are
much greater than the keyspace (for reasonable key sizes),
there isn't an immediate need for a larger table.
3. RC4 gets its security from the fact that the table changes
state over time. The entire table changes every 256 outputs.
WIth a 16-bit table, it would change much slower. For some
attacks, then, it might be reasonable to assume that the table
is a constant. This would make it weaker.
These are just my feelings. I have done no rigerous analysis.
>(And one of these days I'm going to pick up Applied Cryptography, but
>haven't gotten around to it yet. If this particular analysis is right in
>there, please forgive me. :) )
No. There isn't.
Bruce
**********************************************************************
Bruce Schneier, President, Counterpane Systems Phone: 612-823-1098
101 E Minnehaha Parkway, Minneapolis, MN 55419 Fax: 612-823-1590
Free crypto newsletter. See: http://www.counterpane.com
------------------------------
** 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
******************************
