Date: Sat, 29 Dec 2001 19:22:10 -0600 (CST)
From: Tina Bird <[EMAIL PROTECTED]>
Subject: Re: [LIB] Security

DES is a 56-bit key length cipher.  Independent of
implementation, it took 2.1 days for the custom-built
"Deep Crack" machine to test all possible 56-bit keys
in the DES keyspace -- that machine was a multi-processor
system with 10000 CPUs, but balancing the difficulty of
assembling all those processors, the group that did the
work only bought >>33MHz<< processors -- hardly top of
the line even in, uh, 1997/1998 when the work was done.

http://www.cryptome.org/cracking-des.htm

This work was done to force the US government to admit
publicly that its chosen crypto algorithm -- DES -- wasn't
able to protect private data even against a reasonably
well-funded private attacker (the group estimated that, 
given their blueprints and code, which they published
freely, it would cost about $50k US to build a cracker
of your own).  The US government has chosen a replacement
algorithm called Rijndael; commercial implementations 
should be available this year, and the code is already
available.

http://www.nist.gov/aes

All of that being said, there are lots and lots of ways for government
agencies to get access to individual private data --
and yes, Raymond, TEMPEST (which is based on the 
radio emissions that a PC puts out) works on laptops as
well as desktops.  The big problem I have with home users
and individuals not patching their personal machines
isn't the threat to >their< data, it's the high 
likelihood that their machines will be used as
launch points for attacks on other boxes.

tbird -- security geek

On Sat, 29 Dec 2001, Raymond wrote:

> Date: Sun, 30 Dec 2001 09:50:17 +0800
> From: Raymond <[EMAIL PROTECTED]>
> Subject: Re: [LIB] Security
> 
> At 05:42 PM 29/12/2001 -0800, you wrote:
> >Date: Sun, 30 Dec 2001 01:36:17 +0000
> >From: "Matthew Hanson" <[EMAIL PROTECTED]>
> >Subject: Re: [LIB] Security
> >
> >>Date: Sat, 29 Dec 2001 16:35:01 -0800 (PST)
> >>From: David Chien <[EMAIL PROTECTED]>
> >
> >>www.pgpi.com's PGP Disk will help keep your PC secured from outside
> >
> >Something that guy said that threw me was that PGP was an algorithm in itself.  I 
>always thought that PGP was Philip Zimmerman's system of excanging keys... and that 
>he used RSA or some other established algorithm to do the  actual encryption.  My 
>memory is foggy on this.  The only time I used PGPi, I think I recall having to 
>choose an algorithm from a drop-down box.
> 
> IDEA data encryption, RSA for key management, MD5 for hashes. IDEA hasn't been 
>publicly broken yet, I don't think RSA nor MD5 have been sufficiently comprimised to 
>render the PGP system insecure either. IIRC key lengths are also long enough to keep 
>data secure for a decade or 2 barring any major advances in cryptographic techniques 
>(I'm going by what I read in 'Applied Cryptography' first edition here though, I've 
>not read any more recent confirming texts on these algorithms so I could be wrong).
> 
> 
> >Another thing this guy told me was that DES has been broken.
> 
> Well IIRC DES to NSA's public standard can now be broken by brute force anyway given 
>a sufficiently resourced adversary. You could use triple-DES though and make it a lot 
>harder ...
> 
> 
> >  I asked if it's something that would a few super-computers to throw at, and he 
>said no.  That is was a simple high-school hack of some kind.  Like I said... I don't 
>follow this stuff very much... so DES could have patched (can that be done?) by this 
>point for all I know.
> 
> *shrug* did he say DES had been broken or a specific DES implementation has been 
>broken? Bear in mind, there are 2 parts to any piece of encryption software, the 
>algorithm and the implementation. Often the implementation can be broken more easily 
>than the algorithm itself (eg. if the implementation puts a constant header somewhere 
>you can use a known plaintext attack for instance). Of coures, DES itself is getting 
>old now anyway ... IIRC it was based on IBM's Lucifer cipher and supposed to expire 
>in the 90's as a standard anyway (note DES stands for Data Encryption Standard which 
>is pretty ironic I guess).
> 
> 
> >>I laugh as other poor IE/Outlook users keep getting hacked/viruses/exploited
> >>with my Netscape-based, secured Libretto.
> >
> >Can someone explain to me just why Mircosoft thought it would be such a great idea 
>to be able to set an option in Outlook(X?) that would allow email attachments to be 
>launched automatically when an email is opened!!!????  As I recall, some of the first 
>versions came set up that way by default.  
> >Wasn't that what originally inspired the creator of the Love bug?
> 
> Thats Microsoft for you ... remember, security is inversely proportional to 
>convenience. Microsoft's OE just went too far for convenience. If you're paranoid 
>about that, use Pine, good luck any executables getting run there!!
> 
> 
> - Raymond
> 
> ---
> 
> 
> /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\
> |                 | "Does fuzzy logic tickle?"                |
> |   ___           | "My HDD has no reverse. How do I backup?" | 
> |  /__/           +-------------------------------------------|
> | /  \ a y b o t  |          [EMAIL PROTECTED]             |
> |                 |          HTTP://www.raybot.net            |
> | ICQ: 31756092   |   Need help? Visit #Windows98 on DALNet!  |
> \~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~/
> 
> 
> 
> 
> **************************************************************
> http://libretto.basiclink.com - Libretto mailing list
> http://libretto.basiclink.com/archive - Archives
> http://www.picante.com/~gtaylor/portable/faq.html - FAQ
>                  -------TO UNSUBSCRIBE-------
> Reply to any of the list messages. The reply mail should be
> addressed to: [EMAIL PROTECTED] - Then replace any text
> on the message's subject line: cmd:unsubscribe
>               --------TO UNSUBSCRIBE DIGEST------
> Do above but with this on subject line: cmd:unsubscribe digest
> **************************************************************
> 




**************************************************************
http://libretto.basiclink.com - Libretto mailing list
http://libretto.basiclink.com/archive - Archives
http://www.picante.com/~gtaylor/portable/faq.html - FAQ
                 -------TO UNSUBSCRIBE-------
Reply to any of the list messages. The reply mail should be
addressed to: [EMAIL PROTECTED] - Then replace any text
on the message's subject line: cmd:unsubscribe
              --------TO UNSUBSCRIBE DIGEST------
Do above but with this on subject line: cmd:unsubscribe digest
**************************************************************

Reply via email to