Slashdot Mirror
Slashback: Sale, Secrecy, Lasers
That's a lot of Molsen. Bigger R writes contributes this link to a story in the Calgary Herald with more detail on the sale of Corel's Linux division which was mentioned in vague terms the other day. It's going to a company called Xandros, in exchange for cash and equity, so Corel will still have at least some interest in the continued success of Linux, or at least its distro. A snippet: "Xandros Corp. president Michael Bego, who started the Ottawa company recently in preparation for the deal announced Wednesday, is also a shareowner in Linux Global Partners, a privately held New York venture firm that put up $10 million US to start Xandros."
Small, cheap and fun are all good words. An Anonymous Coward writes "There's an announcement of the winners of the embedded linux journal's design contest over at linuxdevices.com. Cool projects -- voting system, digital audio workstation, solar racing vehicle, GizmoCopter Project, and Hacking BigMouth Billy Bass. Follow the urls for the projects which are given in the announcement to learn about each project. Oh, and the prize for winning each category? An all expense paid trip to Costa Rica. Dang, why didn't I enter?"
Stuff that's hard to read. John Sokol writes with an update on the Cayley-Purser Algorithm mentioned here before.
"This story went through some time back about a 16-year-old girl outdoing RSA, but it lacked any discussion of the actual algorithm. This link is her paper that she now has posted on the net. It seems reasonable. Maybe someone here can find a flaw in it?"
Roll 'em. Slide100 writes: "It seems that there is more to the desktop rover that was posted about on Tuesday.
The marketing manager sent me a PDF file that explains some more - apparently, they just don't have the time to update the website.
Each rover comes with a cable that plugs into the transmitter and software to allow control of the rover from your computer (or through TCP/IP).
Additionally, each rover has 'Laser Tag' as an integral part of the vehicle. 10 hits (including sound effects) and your rover is disabled 'till the next match See it here. BTW - I have nothing to do with the company, I just think its very cool."
Obviously the poster did not read the article himself. In a postscript of her own paper (http://www.cayley-purser.ie/#Post_Script__An_Atta ck_on_the_CP_Algorit), a successful attack is described. The CP algorithm is only useful as a private key scheme, which is no big news.
You know this is a really cool little deal. What is really going to be pretty neat is when you can set up rock'em socke'em robots.
With lasers on them to plug into your computer though a wireless network work. Then play with them though a head mounted camera! -- Opss wait that is Quake.
Neck_of_the_Woods
#/usr/local/surf/glassy/overhead
I don't see anything about laser tag on the vehicle, but I did get this when I was trying to find it:
This page is being modified, Sorry about that -
come back later and we should have the
changes completed. Please Click your
browser's "Back" Button.
Thank You!
Heh. I guess maybe they're trying to add it, or trying to keep from getting slashdotted too badly?
Umm, she has pointed out the flaw in her own post-script. If I read it correctly, it says
that if you know the public key and one encrypted
message then you can break it.
There is a book entitled "In Code" by Sarah and her father, published by Workman, NY, that very interestingly describes her work. In it she tells of being advised by a well known mathematician that there is an attack on the method she described, due to the famous Caley-Hamilton theorem.
Soposidly the encryption is 20 times faster than RSA. So if it's deemed as secure as RSA, she recommends switching over to it. I could be an idiot, but wouldn't it be faster to brute force crack it if it is 20 times speedier than RSA?
1337
What I find most interesting is the use of an inverse posiform discombobulator to repel sophisticated man in the middle attacks. Combined with the infix digestive emulator, this will be a key component in the "Digital Nervous System" of the future.
Superficially, the Algorithm appears similary to the Bosun-Smee cipher which has been shown to be vulnerable to chosen plaintext attacks when not run in output feedback mode. Personally I think an approach based on a 4x4 Holmes-Longpole network would have been a preferable starting point, but this would be more vulnerable to differential polyp-cystizing cryptanalyis. This can (theoretically) be compensated for by a field-coit gestation transform, but no one is quite sure how to overcome the slow encryption speed of such a system.
All in all it's a fairly interesting approach and I intend to study it further. I'd love to hear anyone elses opinion though.
--Shoeboy
It's very strange that I am moved to post by this thing.
Many of the most useful little devices started as toys. For instance the Aztec and Inca civilizations had wheels on their childrens toys but not on their wagons and carts.
This little toy, or at least the model of a cheap travelling rover based on existing tech, could lead to cheaply revolutionized communications in cities and other compact spaces, such as aircraft carriers, etc.
Using these devices as messengers would cut down on traffic and save gasoline and car-use resources in many cities.
It sounds weird to us not because it's scientifically or fiscally implausible but because we are involved in a car culture.
Global warming is good for you!
if you want quickies, do what i do-- go to memepool.
Humpty Dumpty was pushed.
...codes. Controversial website adequacy.org has the skinny on autistic people being used by the NSA to crack RC5. Apparantly, each autist is capable of 1 megaflop per second, and there are many thousands of unused autists in our fine country. Are we not using their potential as we should?
KTB:Lover, Poet, Artiste, Aesthete, Programmer.
There is no
My crypto (and linear algebra) is somewhat rusty, it's been a few years, but I think that the problem is that Alice can't sign with her secret code. The paper mixes Alice and Bob a little, but assuming that Alice publishes and Bob encrypts in the algorithm, then unlike the RSA keys, Alice can't sign a message with her private key for all the world to see (using her public key). That is a big advantage of RSA.
Is it a feature or a bug? what's the difference?
Ummm...I can't seem to get the lazer game...any ideas? Appearently the page is up for construction.
The anti-salmon
The algorithm has already been shown to have fatal flaws (search for Plaintext Defense). Sarah Flannery herself was quoted as such. However, there is nothing to say that the flaw cannot be eliminated in the next version. To say the least, this algorithm provides a strong foundation on which to build.
I just skimmed over the paper, but her algorithm seems a lot like RSA except that she encrypts/decrypts with pre computed matricies instead of exponentiations. There are some easy ways to speed exponentiation up when doing it modulo... I wonder if she took that into account when comparing.
It appears that the matricies that are used as keys instead of the normal RSA keys are the cause of the crack. They give too much extra info to a possible attacker.
I want to look this over better when I have more time, it's pretty interesting. God knows I didn't understand RSA that well when I was 16.
garc
Controversial website adequacy.org has the skinny on autistic people being used by the NSA to crack RC5. Apparantly, each autist is capable of 1 megaflop per second, and there are many thousands of unused autists in our fine country. Are we not using their potential as we should?
Of course we're not! Can you imagine a beowulf cluster of those things?!
NO CARRIER
If I were hacking Big Mouth Billy Bass to put Linux on it, I'd replace that annoying "Take Me To The River" sample with the classic "Hello, this is Linus Torvalds, and I pronounce Linux as Leenux."
Of course, that would probably get old really fast, just like the standard sample. Plus, no one really wants to hear Linus say the same thing every day (except maybe his kids, who want to hear the "scary module under the bed" story every night).
I'd really like to replace the current sample with the Talking Heads cover of the same song. Much better. Maybe give Big Mouth Billy a Big David Byrne Suit too.
"I may not have morals, but I have standards."
You're counting cards?
We're counting cards....
Uhh...huh...
Think about it, get one of those desk rovers, strap an x10 camera to it, drop it in somewhere interesting, like the runways in your office for cabling and in the ceiling... Instant remote control first person shooter!
The right approach is to decompress ontlogogical entropism. You have to do this, or all you're left with is a lot of noise!
i read that link. It doesnt say ANYTHING about NSA using autistic people, nor autistics being able to do computations at 1 megaflop/second. Besides, theose ppl cant be trusted as autistics are easily mislead and such, so anything they do/see can be easily weasled out of them. Learn to read, then try making false claims.
The One Rule Of Chess You'll Ever Need: Don't play someone who carries a kit in their bookbag.
I don't know a lot about alternative encryption schemes, but the relatively simple math involved suggests that this has the potential to be fast by any standard.
I'm wondering if it might be worthwhile to do a mixed scheme whereby E is returned using RSA or some other technique and the bulk of the message is encrypted using C-P. If I read the comments correctly one has to retrieve the unique E for each message in order to break the system and read the encrypted text of that message.
The table on the end show that a 250k file takes 50 minutes to encrypt/decrypt with RSA on the P 133 under Mathematica (relatively inefficient), and only a little more than 2 mins for C-P. Even given faster computers and optimized code, it would seem that one might see significant gains in speed if you are encrypting multi-megabyte files.
Any thoughts?
Block ciphers like DES and AES are much faster to compute than even C-P, since they don't require multiprecision arithmetic. AES, in particular, screams.
but will the cluster run linux?
Shift happens. Fire it up.
Block ciphers like DES and AES are much faster to compute than even C-P, since they don't require multiprecision arithmetic. AES, in particular, screams.
The encryption/decrytpion of C-P uses only matrices of integers (all operations are modulo n). Having another encryption method might not be needed, but you're objection doesn't seem accurate.
Or is there some meaning for "multiprecision arithmetic" other than multiprecision floating-point arithmetic?
Nobody uses RSA to encrypt multi-megabyte files anyway. Traditionally you use a public key algorythim to encrypt a session key (realistically no bigger than 32 bytes or so), and then use the session key to encrypt the message with a well known and tested symmetric algorythim (like 3Des, Blowfish, IDEA, etc).
-- The act of censorship is always worse than whatever is being censored. Always.
When does this chick turn 18? I'm in love....
Well yes, but only with x86 emulation ... damn binary-only drivers.
I have a couple of questions. hopefully someone
who knows a lot about encryption can answer.
The paper claims there are two lines of attack:
1. B={(AC)^-1}*C
2. G=C^r
For the first attack to be hard we need to
pick p and q carefully. Does this restriction
affect the ease of the second attack?
Also, this negligible probability that the first
attack will be easy seems to have a PR disaster
written all over it. Does RSA have similar
probability issues?
Will it still be called corel linux or xandros linux?
===> An eye for an eye makes everyone blind - MG
Imagine a beowulf cluster of those slobbering bastards!
The current Slashdot moderation system is made by gay communists!
--
Dave
Anyway - factor of 20 is negligible when we I talking about attacks - but not negligible, when we are talking about usage.
<^>_<(ô ô)>_<^>
Surely RSA's only used to set up the session key for ssh? Once you're connected, you're using a symetrical cryptography algortihm. So C-P could only save you time in the connection setup.
--
TBBle
Paul "TBBle" Hampson
Paul.Hampson@Pobox.Com
This site is cool; where is the babe who has an hour to spare?
- Kaos games and encryption systems developer
It's not even a new update to old news.
you're not very bright are you?
Everyone-- me, the old lady next door, the kid down the street, the dog-- knows that field-coit gestation transforms and ontological decompression over a finite field are about as secure as a tipsy girl's chastity on prom night. Field-coit, when push comes to shove, is hardly more than a complex-- but certainly tractable-- hex-stacked XORing from an arbitrary (but by no means random) set of figures of undefined length. And ontological demopression?!? Christalmighty! Not only is it slow (maybe you can wait 1345 months to encrypt "The Old Man in the Sea", but I myself have a wife and dog to feed), but you actually end up with LESS entropy than using, say, triple-DES with unique keys-- which also has the tiny-tiny-twee little advantage of not taking 112 YEARS!!!
This is what I hate about this ENTIRE FIELD: Some gintch in Ireland comes up with a cute (if entirely infeasable)cryptosystem (which, I might add, she has already aknowledged and published the weaknesses of!), and then I have to listen to every blowhard from here to Katamandu go off about how great field-coit and ontological decompression are. Christ, it's wors than talking about laptops with Mac-Addicts!
Some days, I'd rather be flipping burgers.
Much Love,
"S"HM
*****
(I refuse to spellcheck out of contempt for your belief system)
I wonder, if since the desktop rover can be hooked up via tcp/ip to a computer, and since it has laser tag capabilities...
Could one hack it to be a remote IrDA port? maybe be used as a second remote control for your TV? There must be some super-cool use for a radio controlled remote control...
At least you'd never lose it!!
"He's more machine now than man, twisted and evil."
The government has done tons of wacky stuff...e.g., the CIA had (or maybe still has??) a program, to train people to psychicly detect the presence of russian submarines. The scary thing is that this was apparently really working (or else why would the spooks care?).
It's 10 PM. Do you know if you're un-American?
I think you need a little anger management. #1, it was strapping the laser to the head of the robot, which is controlled by the computer. Hence the quake.....You know robot like men running rouand on your computer with lasers shooting at each other. #2 Breath good air in, Breath bad air out, no hum a C# note and think "life is good". Maybe that will help.
Neck_of_the_Woods
#/usr/local/surf/glassy/overhead
Wouldnt that then be large scale arithmetic rather than large precision arithmetic?
Just curious.
Video meliora proboque deteriora sequor - Ovidius
what exactly would you be getting at? If you were to follow the link the post made, you wont see references to NSA nor how well autistic people can do math.
The One Rule Of Chess You'll Ever Need: Don't play someone who carries a kit in their bookbag.
if you reverse the polarity of the neutron flow.
A cluster of which? The sixteen year old girls or the autistic people? Personally, I'd prefer a cluster of 24 year old ladies, but I doubt I'd ever get that wish.
// file: mice.h
#include "frickin_lasers.h"