Cryptography-Digest Digest #191, Volume #13 Mon, 20 Nov 00 11:13:01 EST
Contents:
How to hash a 50 MB byte file? (Cheng-Fen)
Long Division Algorithm ([EMAIL PROTECTED])
Question regarding OS's. ("Juri")
consecuitive Fibonacci Numbers ([EMAIL PROTECTED])
Re: Long Division Algorithm (John Savard)
Re: consecuitive Fibonacci Numbers (Paul Rubin)
Re: consecuitive Fibonacci Numbers ("Douglas A. Gwyn")
Re: Long Division Algorithm ("Douglas A. Gwyn")
Re: Question regarding OS's. ("Douglas A. Gwyn")
Re: Rijndael Key Schedule (Mok-Kong Shen)
Re: The SHAs (Bob Deblier)
Re: Cryptogram Newsletter is off the wall? (Daniel James)
Re: Cryptogram Newsletter is off the wall? (Daniel James)
Re: consecuitive Fibonacci Numbers ("Charles Matthews")
Re: Question regarding OS's. (Dido Sevilla)
Re: Cryptogram Newsletter is off the wall? (Matthias Murra)
Re: Cryptogram Newsletter is off the wall? ([EMAIL PROTECTED])
Trusted/Untrusted components...Cryptogram Newsletter is off the wall?
([EMAIL PROTECTED])
Re: Cryptogram Newsletter is off the wall? (Niklas Frykholm)
Total $ spent on voice encryption ([EMAIL PROTECTED])
Re: How to hash a 50 MB byte file? (Tom St Denis)
Re: Total $ spent on voice encryption (Jeffrey Williams)
Re: Long Division Algorithm (Charles Blair)
Re: Cryptogram Newsletter is off the wall? ([EMAIL PROTECTED])
Re: simple proof of summation (Bob Silverman)
Re: [Question] Generation of random keys (Alan Rouse)
simple proof ([EMAIL PROTECTED])
Re: simple proof (Paul Crowley)
Re: Long Division Algorithm (Arturo)
----------------------------------------------------------------------------
From: Cheng-Fen <[EMAIL PROTECTED]>
Subject: How to hash a 50 MB byte file?
Date: 20 Nov 2000 03:37:39 GMT
Which Algo is faster?
Or are there any on-the-shelf product?
Thanks
------------------------------
From: [EMAIL PROTECTED]
Subject: Long Division Algorithm
Date: Mon, 20 Nov 2000 04:34:31 GMT
By using the long division algorithm to divide (1-x) into 1 how can we
show that
1/(1-x) = 1+x+x^2+x^3+...
Since this is an assignment, any suggestions will defenately
appreciated.
Thank you for your co-operation
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: "Juri" <[EMAIL PROTECTED]>
Subject: Question regarding OS's.
Date: Mon, 20 Nov 2000 04:55:57 GMT
Hello,
I am just curious, why OS do you, cryptographers, use?
Windows, Linux, Unix or something else?
If Unix, what distributor of Unix? Thanks very much.
Juri
------------------------------
From: [EMAIL PROTECTED]
Subject: consecuitive Fibonacci Numbers
Date: Mon, 20 Nov 2000 05:05:23 GMT
How can we show that every two consecuitive Fibonacci Numbers are
coprime.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (John Savard)
Subject: Re: Long Division Algorithm
Date: Mon, 20 Nov 2000 05:15:07 GMT
On Mon, 20 Nov 2000 04:34:31 GMT, [EMAIL PROTECTED] wrote, in part:
>By using the long division algorithm to divide (1-x) into 1 how can we
>show that
>1/(1-x) = 1+x+x^2+x^3+...
>Since this is an assignment, any suggestions will defenately
>appreciated.
>Thank you for your co-operation
Well, you should do your own homework. Think "binomial theorem".
After all, 1/9 = .11111111.... which is almost an example of that
(just change to 1/(1-1/10) = 1.111111...).
John Savard
http://home.ecn.ab.ca/~jsavard/crypto.htm
------------------------------
From: Paul Rubin <[EMAIL PROTECTED]>
Subject: Re: consecuitive Fibonacci Numbers
Date: 19 Nov 2000 21:47:58 -0800
H
[EMAIL PROTECTED] writes:
> ow can we show that every two consecuitive Fibonacci Numbers are
> coprime.
Do your own homework...
------------------------------
From: "Douglas A. Gwyn" <[EMAIL PROTECTED]>
Subject: Re: consecuitive Fibonacci Numbers
Date: Mon, 20 Nov 2000 01:56:40 -0500
Paul Rubin wrote:
> [EMAIL PROTECTED] writes:
>> How can we show that every two consecuitive Fibonacci Numbers are
>> coprime.
> Do your own homework...
Yeah, but a more helpful answer is "how does one
show some property holds for an entire sequence,
Grasshopper?"
------------------------------
From: "Douglas A. Gwyn" <[EMAIL PROTECTED]>
Subject: Re: Long Division Algorithm
Date: Mon, 20 Nov 2000 02:06:01 -0500
John Savard wrote:
> On Mon, 20 Nov 2000 04:34:31 GMT, [EMAIL PROTECTED] wrote, in part:
>> By using the long division algorithm to divide (1-x) into 1 how can we
>> show that
>> 1/(1-x) = 1+x+x^2+x^3+...
>> Since this is an assignment, any suggestions will defenately
>> appreciated.
> Well, you should do your own homework. Think "binomial theorem".
But that isn't explicitly using the division algorithm.
This exercise is one that practically works itself once
one attempts it. The only "trick" is in seeing how to
relate the general step to the first step, but it's
easy to see.
------------------------------
From: "Douglas A. Gwyn" <[EMAIL PROTECTED]>
Subject: Re: Question regarding OS's.
Date: Mon, 20 Nov 2000 02:10:34 -0500
Juri wrote:
> I am just curious, why OS do you, cryptographers, use?
For cryptanalysis, the UNIX shell environment
(filters & pipelines) is very convenient.
These days a similar environment is available
for most OSes, although sometimes not "out of
the box", i.e. you have to install it yourself.
------------------------------
From: Mok-Kong Shen <[EMAIL PROTECTED]>
Subject: Re: Rijndael Key Schedule
Date: Mon, 20 Nov 2000 11:35:03 +0100
John Myre wrote:
>
> Trish Conway wrote:
> <snip>
> > And why have the 4 other AES finalist algorithms ensured that the subkeys
> > are derived from the userkey in a more complex fashion?
> <snip>
>
> This is a good question, and I hope somebody competent will
> address it. I don't know the answer. Don't forget, the other
> AES finalists would certainly not have bothered to make their
> key scheduling longer if they weren't sure they needed it.
>
> In general, it is certainly true that you have to look at a
> cipher in its entirety: you can't conclude insecurity from
> a single element. So a good answer to this question would
> indicate, in a general way, how the structure of Rijndael is
> expected to remove the need for complex key scheduling, as
> compared to, say, Twofish.
I also like very much to know the principle of design of
keys cheduling, which seems to be little treated in the
literature. A guess of mine is that a good key scheduling
makes the relationship between the round keys themselves
complicated enough so that the analyst has difficulty to
exploit that. But this statement is certainly very vague.
M. K. Shen
------------------------------
From: Bob Deblier <[EMAIL PROTECTED]>
Subject: Re: The SHAs
Date: Mon, 20 Nov 2000 11:45:22 +0100
kihdip wrote:
> Thanks,
>
> Although: Is it Mbytes or Mbits ?
>
> Kim
That's MB(ytes) versus Mb(its). AFAIK these are the standard
abbreviations.
Sincerely
Bob Deblier
Virtual Unlimited
------------------------------
From: Daniel James <[EMAIL PROTECTED]>
Subject: Re: Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 10:56:22 GMT
Reply-To: [EMAIL PROTECTED]
In article <kyTR5.977$17.26932@stones>, Brian Gladman wrote:
> The issue here is not that of the best computer environments we can think of
> for making digital signatures but rather that of the security we can expect
> from typical computer environments in which such signatures will be made.
There are actually two issues here. One, from the signer's perspective, is
that you cannot generate a signature securely on an insecure system; the
other, from the recipient's perspective, is that you cannot trust a signature
in a received message unless you know not only that your own system is secure
but also that the signer's system was secure.
Knowing that a signature was generated in a secure manner when you have no
knowledge of the signing system is a much harder task than ensuring that your
own system has not been compromised by a Trojan (which, admittedly, is not
exactly an easy task either).
Cheers,
Daniel.
------------------------------
From: Daniel James <[EMAIL PROTECTED]>
Subject: Re: Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 10:56:22 GMT
Reply-To: [EMAIL PROTECTED]
In article <8v94dv$ljj$[EMAIL PROTECTED]>, wrote:
> ... you dont have to make the DS on the computer being open to abuse
> and intrusion.
>
> Why not make the DS on a smart card, e.g a Javacard....
>
> The Keys never leave the card, and you can write the whole DS process
> on the card....this is now feasible with the high end 32 bit Javacards.
Sure, there are smartcards that can compute digital signatures - you can
get dedicated DS cards that do this, and they don't have to be very
high-end, you certainly don't need to go to JavaCard - and they do store
the key securely.
There are still attacks that can compromise the use of those keys. I
could write a Trojan that would sit on your system waiting for you to use
your smartcard to sign something and then, after you have activated the
card by entering your PIN, I could switch messages so that you sign
something other than what you intended. As far as the smartcard is
concerned you asked it to compute a signature, you provided a valid PIN,
and it generated a signature.
Use of smartcards certainly helps to make a system secure, but it doesn't
cover all the loopholes.
Cheers,
Daniel.
------------------------------
From: "Charles Matthews" <[EMAIL PROTECTED]>
Subject: Re: consecuitive Fibonacci Numbers
Date: Mon, 20 Nov 2000 11:39:06 -0000
Douglas A. Gwyn wrote
>Yeah, but a more helpful answer is "how does one
>show some property holds for an entire sequence,
>Grasshopper?"
That would be induction. Which works in this case ...
Charles
------------------------------
From: Dido Sevilla <[EMAIL PROTECTED]>
Subject: Re: Question regarding OS's.
Date: Mon, 20 Nov 2000 20:15:42 +0800
Juri wrote:
>
> Hello,
> I am just curious, why OS do you, cryptographers, use?
> Windows, Linux, Unix or something else?
> If Unix, what distributor of Unix? Thanks very much.
> Juri
I personally use Linux. And I think Vincent Rijmen loves Linux about as
much as I do. :-) lolz... I flatter myself to call myself a
cryptographer. But well, I guess there's a great appeal for open
systems like Linux for cryptographers and the similarly paranoid. Less
possibility of backdoors and security risks. Nobody on this newsgroup
believes in security by obscurity, I think...
--
Rafael R. Sevilla <[EMAIL PROTECTED]> +63 (2) 4342217
ICSM-F Development Team, UP Diliman +63 (917) 4458925
OpenPGP Key ID: 0x0E8CE481
------------------------------
From: Matthias Murra <[EMAIL PROTECTED]>
Subject: Re: Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 13:32:39 +0100
Bruce Schneier wrote:
>
> We've reached the point where passwords do not provide security
> against off-line attacks.
>
> There is an upper limit of what people can be reasonably expected to
> remember and type in. And over the years, the efficacy of dictionary
> attacks has increased. A few years ago, the two crossed.
This doesn't necessarily have to be true -- see the following paper from
Ross Anderson's home page:
http://www.cl.cam.ac.uk/ftp/users/rja14/tr500.pdf
--
"Opinions expressed are my own. No one in their right mind would
claim them." (Origin unknown)
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 12:25:51 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> In article <8v94dv$ljj$[EMAIL PROTECTED]>, wrote:
> > ... you dont have to make the DS on the computer being open to
abuse
> > and intrusion.
> >
> > Why not make the DS on a smart card, e.g a Javacard....
> >
> > The Keys never leave the card, and you can write the whole DS
process
> > on the card....this is now feasible with the high end 32 bit
Javacards.
>
> Sure, there are smartcards that can compute digital signatures - you
can
> get dedicated DS cards that do this, and they don't have to be very
> high-end, you certainly don't need to go to JavaCard - and they do
store
> the key securely.
>
> There are still attacks that can compromise the use of those keys. I
> could write a Trojan that would sit on your system waiting for you to
use
> your smartcard to sign something and then, after you have activated
the
> card by entering your PIN, I could switch messages so that you sign
> something other than what you intended. As far as the smartcard is
> concerned you asked it to compute a signature, you provided a valid
PIN,
> and it generated a signature.
>
> Use of smartcards certainly helps to make a system secure, but it
doesn't
> cover all the loopholes.
>
> Cheers,
> Daniel.
>
>
That depends on how your application is written. If there are security
features in your application, which says something like:
OK This message is designated to go into the smart card...
Put this message in a designated and secure are of memory
OK..Now I sign and send this data to the smart card
OK Now I get signed data from the smart cards
Has an intruder changed the data in between?
OOPS the data has been changed..
Trigger an alarm...
and so on.....I guess this is an internal IPsec between the smart card
reader and the "External System"
That is just one idea...
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED]
Subject: Trusted/Untrusted components...Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 12:33:32 GMT
In the thread below ..Cryptogram Newsletter is off the wall?..the
argument presented by Bruce Schneier is that because a computer is not
a Trusted component, we dont know if we can Trust a DS ..
There is no such thing as a Trusted PC? No
Can you Trust the OS? No...There is no Trusted Secure OS...
Can you Trust External Threates? No
Can you Trust the Browser? No
Can you Trust your Server?
We have to do the best with the components we have...A Trusted System
is an idealised system like the OTP....
So we either give up or build systems with the best secuirty options
designed in...
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (Niklas Frykholm)
Subject: Re: Cryptogram Newsletter is off the wall?
Date: 20 Nov 2000 09:11:02 GMT
> (This is a simplification. With a few exceptions, you
>can't take a signed document into court and argue that Alice signed
>it. You have to get Alice to testify that she signed it, or bring
>handwriting experts in and then it's your word against hers. That's
>why notarized signatures are used in many circumstances.)
Perhaps we will see something similar to notarized signatures for
digital signatures as well. The notary public would provide a secure
computer where you could bring your document and signing card. The
notary public would be responsible for the security of the signing
station as well as for verifying your identity.
I think it is important to realize that for security, we must look
not only at the signature (whether digital or handwritten) but at
the entire signing situation.
Do handwritten singature prevent forgeries --- probably not. Someone
can imitate my signature (albeit, perhaps not good enough to fool
an expert) or my singature could be lifted from one document to
another. I've frequently signed multi-page documents only on the last
page --- how do I know that no one exchanges the remaining pages?
Or someone could force me to sign the document.
Do handwritten signatures provide undeniability --- definitely not.
I have signed my credit card receipts Haile Selassie and Donald
Duck --- nobody seems to care. Since signature
experts are not called in until there is a trial I could have a
friend forge my singature on a document and later deny that
I've signed it --- the experts would prove me right.
So, handwritten or digital, it is the signing situation that
matters. We might be able to do away with handwritten signatures,
but I do not think we will be able to do away with the notary
publics.
// Niklas
------------------------------
From: [EMAIL PROTECTED]
Subject: Total $ spent on voice encryption
Date: Mon, 20 Nov 2000 12:48:12 GMT
Why do business spent Billions $ for all kinds of data security but will not
spend money on voice encryption as if wiretapping and eavesdropping is not
happening.
Could somebody advise?
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Tom St Denis <[EMAIL PROTECTED]>
Subject: Re: How to hash a 50 MB byte file?
Date: Mon, 20 Nov 2000 13:05:07 GMT
In article <8va6a3$sfg$[EMAIL PROTECTED]>,
Cheng-Fen <[EMAIL PROTECTED]> wrote:
> Which Algo is faster?
> Or are there any on-the-shelf product?
Hmm most competent implementations of SHA1 will be able to hash it in a
few seconds.
Just look up the source!
Tom
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Jeffrey Williams <[EMAIL PROTECTED]>
Subject: Re: Total $ spent on voice encryption
Date: Mon, 20 Nov 2000 07:23:18 -0600
Well, there is almost certainly a myiad of reasons. However, I would suspect
that the most prominent reason is likely that the people controlling the dollars
know a lot about business but very little about computers, data, or security.
You do not have to understand technology to use it (actually, it seems like the
better the implementation, the less the user must understand the underlying
technology).
Jeff
[EMAIL PROTECTED] wrote:
> Why do business spent Billions $ for all kinds of data security but will not
> spend money on voice encryption as if wiretapping and eavesdropping is not
> happening.
>
> Could somebody advise?
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
------------------------------
Subject: Re: Long Division Algorithm
From: [EMAIL PROTECTED] (Charles Blair)
Date: Mon, 20 Nov 2000 14:42:28 GMT
>>> By using the long division algorithm to divide (1-x) into 1 how can we
>>> show that
>>> 1/(1-x) = 1+x+x^2+x^3+...
>>> Since this is an assignment, any suggestions will defenately
>>> appreciated.
The infinite series diverges for some values of x. You should
look up theorems about multiplication of infinite series. Or
``geometric series.''
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Cryptogram Newsletter is off the wall?
Date: Mon, 20 Nov 2000 15:05:10 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> In article <8v94dv$ljj$[EMAIL PROTECTED]>, wrote:
> > ... you dont have to make the DS on the computer being open to abuse
> > and intrusion.
> >
> > Why not make the DS on a smart card, e.g a Javacard....
> >
> > The Keys never leave the card, and you can write the whole DS
process
> > on the card....this is now feasible with the high end 32 bit
Javacards.
>
> Sure, there are smartcards that can compute digital signatures - you
can
> get dedicated DS cards that do this, and they don't have to be very
> high-end, you certainly don't need to go to JavaCard - and they do
store
> the key securely.
>
> There are still attacks that can compromise the use of those keys. I
> could write a Trojan that would sit on your system waiting for you to
use
> your smartcard to sign something and then, after you have activated
the
> card by entering your PIN, I could switch messages so that you sign
> something other than what you intended. As far as the smartcard is
> concerned you asked it to compute a signature, you provided a valid
PIN,
> and it generated a signature.
>
> Use of smartcards certainly helps to make a system secure, but it
doesn't
> cover all the loopholes.
For online transactions you could have a smartcard with a keypad that
created a secure session through the terminal/ATM/whatever to a
remote server that stored the private key using the a PIN and a protocol
like SNAKE (plug plug :-), SPEKE, or SRP. The document/data could
then be sent to the remote server for signing either via the card, or
directly from the terminal/ATM/whatever with a MAC generated by the
card. Its even feasible that the card could have a little LCD display
so that you could visibly verify the document, data or content hash.
This would avoid private keys being sent all over the place, and you
dont have to trust the software to do what it says. All you trust is
the your own personal smartcard which doesnt even need to
store your private key (I guess you might want it to for offline
transactions tho).
ttfn
PG.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Bob Silverman <[EMAIL PROTECTED]>
Subject: Re: simple proof of summation
Date: Mon, 20 Nov 2000 15:17:57 GMT
In article <8va37a$ckb$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> OK,
> As an assignemnt I am asked to solve the following Equation
> Sum{i =1,n}Sum{j= 1,i}(i(i-1)/2+j-1)
I see no equation here.
>
> OK
>
> As I was asked to find the solution for it.
As there is no equation, there can be no solution.
Perhaps you would like to restate your problem?
If what you are asking for is the sum expressed as an
algebraic expression in n, you did not ask for this.
Hint: Try Induction
--
Bob Silverman
"You can lead a horse's ass to knowledge, but you can't make him think"
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Alan Rouse <[EMAIL PROTECTED]>
Subject: Re: [Question] Generation of random keys
Date: Mon, 20 Nov 2000 15:28:39 GMT
The hard part is knowing how much entropy you have collected. Perhaps
you want to create a 160-bit key, so you need 160 bits of entropy. How
can you be certain that you have that much? Tough question.
Yes you can be "very conservative" and collect 2x, or 10x, or 100x as
much entropy-containing data as you think you need. But you still
don't KNOW how much you have. Perhaps the entire sample is completely
determined by a relatively small number of controlling factors. This
is a difficult problem that cannot be solved by software. It requires
careful study on a case-by-case basis. And even then you won't KNOW.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED]
Subject: simple proof
Date: Mon, 20 Nov 2000 15:27:10 GMT
Let's say that we have a function, such that
f(n) = Sum{i =0, n-1} 2^i
how can we show these two:
one:
f(n) = f(n-1) +2^(n-1)
second:
f(n) = 2f(n-1) +1
any suggestion for a good start
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Paul Crowley <[EMAIL PROTECTED]>
Subject: Re: simple proof
Date: Mon, 20 Nov 2000 16:02:07 GMT
[EMAIL PROTECTED] wrote:
>
> Let's say that we have a function, such that
>
> f(n) = Sum{i =0, n-1} 2^i
>
> how can we show these two:
> one:
> f(n) = f(n-1) +2^(n-1)
If you can't solve this first one, take a different subject.
--
__
\/ o\ [EMAIL PROTECTED]
/\__/ http://www.cluefactory.org.uk/paul/
------------------------------
From: Arturo <[EMAIL PROTECTED]=NOSPAM>
Subject: Re: Long Division Algorithm
Date: Mon, 20 Nov 2000 16:36:01 +0100
On Mon, 20 Nov 2000 14:42:28 GMT, [EMAIL PROTECTED] (Charles Blair) wrote:
>>>> By using the long division algorithm to divide (1-x) into 1 how can we
>>>> show that
>>>> 1/(1-x) = 1+x+x^2+x^3+...
>>>> Since this is an assignment, any suggestions will defenately
>>>> appreciated.
>
> The infinite series diverges for some values of x. You should
>look up theorems about multiplication of infinite series. Or
>``geometric series.''
It can be shown via the so-called Taylor series. If you have a function
F(x), it can be proven that for some a the following holds:
F(x) = F(a) +f�(a)*(x-1) + f��(a)(x-a)^2 / 2 ... f���� (n-1 �s) (a)*(x-a)^(n-1)
/ (n-1)! + Rn
where the primes indicate derivation of the function F, and Rn is the
remainder. If you make a=0 (MacLaurin series) and F(x)=1/(1-x), you have the
division algorithm above. But the series 1+x+x^2 ... only converges (it can
only be sumed) when the absolute value of x is less than one.
Be warned, however
------------------------------
** 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
******************************
