Looking to help with a crypto design

Looking to help with a crypto design

[snagnbytz]

Well i am not much on posting in GoV related newsgroups or mailing list but i
have checked some things out with SElinux and i agree that the best way to run
a secure system is at the kernel level of encryption.


I have been working on a color based encryption for over a year and half now
which deals with all 16.7 million colors and the level in which i built it is
unique because its combinations is 16.7million to the power of 16.7 million by
using 255of reds greens and blues you can make color filters or take something
similiar to building a raster interrupt with reading and writing the encryption
back to itself which i have mapped out ways to make this possible i think that
using a color based encryption would work on more then just this such as
fiber-optic with different colors only able to respond to the same color it
generates. I do alot of research on my own if i could get back to college i
might be able to make this color based encryption work with more then just
something like what i am seeing here with SElinux.

I will try to see if i can build something of this nature into my own kernel
down the road til then i need a better job so i can get back to college. 

SnaGnBytZ

--
You have received this message because you are subscribed to the selinux list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.

Re: Looking to help with a crypto design

[Dustin Reyes]

Would it be good form to laugh at this point?

-Dustin

On Sun, Feb 11, 2001 at 08:05:22PM -0500, snagnbytz wrote:
> Well i am not much on posting in GoV related newsgroups or mailing list but i
> have checked some things out with SElinux and i agree that the best way to run
> a secure system is at the kernel level of encryption.
> 
> 
> I have been working on a color based encryption for over a year and half now
> which deals with all 16.7 million colors and the level in which i built it is
> unique because its combinations is 16.7million to the power of 16.7 million by
> using 255of reds greens and blues you can make color filters or take something
> similiar to building a raster interrupt with reading and writing the encryption
> back to itself which i have mapped out ways to make this possible i think that
> using a color based encryption would work on more then just this such as
> fiber-optic with different colors only able to respond to the same color it
> generates. I do alot of research on my own if i could get back to college i
> might be able to make this color based encryption work with more then just
> something like what i am seeing here with SElinux.
> 
> I will try to see if i can build something of this nature into my own kernel
> down the road til then i need a better job so i can get back to college. 
> 
> SnaGnBytZ
> 
> --
> You have received this message because you are subscribed to the selinux list.
> If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
> the words "unsubscribe selinux" without quotes as the message.

-- 
---------------------------------------------------------------
Dustin Reyes - crusader@linuxgames.com "Freedom is the right of
LinuxGames - http://www.linuxgames.com   all sentient beings"
---------------------------------------------------------------

--
You have received this message because you are subscribed to the selinux list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.

Re: Looking to help with a crypto design

[Johnathon Day]

On Sun, 11 Feb 2001, Dustin Reyes wrote:

> Would it be good form to laugh at this point?

I dunno. Have you put the form through lint?
 
> -Dustin
> 
> On Sun, Feb 11, 2001 at 08:05:22PM -0500, snagnbytz wrote:
> > Well i am not much on posting in GoV related newsgroups or mailing list but i
> > have checked some things out with SElinux and i agree that the best way to run
> > a secure system is at the kernel level of encryption

English Yoda School of closed, is it.

> > I have been working on a color based encryption for over a year and half now
> > which deals with all 16.7 million colors and the level in which i built it is
> > unique because its combinations is 16.7million to the power of 16.7 million by
> > using 255of reds greens and blues you can make color filters or take something
> > similiar to building a raster interrupt with reading and writing the encryption
> > back to itself which i have mapped out ways to make this possible i think that
> > using a color based encryption would work on more then just this such as
> > fiber-optic with different colors only able to respond to the same color it
> > generates. I do alot of research on my own if i could get back to college i
> > might be able to make this color based encryption work with more then just
> > something like what i am seeing here with SElinux.

Security != Encryption
Colour != (Security | Encryption)

(Well, unless you're a tetrachromat, of course, in which case selectively
invisible ink becomes a real possibility.)

As for the description given, I would strongly urge the author to apply
for a top position at Microsoft, where they will fit right in. Indeed, I
have no doubt whatsoever that they could improve the security there.

Simply put, encryption is the translation of one data stream into a second
data stream, in a manner that makes the inverse operation non-trivial for
a third-party. It really doesn't make any difference if you're using
colours, bitstreams, or giant abacusses strapped to the back of elephants.

The translation function is all that really matters. And the challange is
to develop one that does not reduce to something much simpler -AND- one
that does not produce the correct decrypted stream for multiple
keys. This is NOT trivial. If it was, Julius Caesar's cypher would still
be in use, because there'd be no need for anything better.

Jonathan Day



--
You have received this message because you are subscribed to the selinux list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.

Re: Looking to help with a crypto design

[Sandy Harris]

snagnbytz wrote:
> 
> Well i am not much on posting in GoV related newsgroups or mailing list ...

> I have been working on a color based encryption ...

Methinks this question would be better asked on the sci.crypt newsgroup than here.
The security enhancements this group is adding to Linux are changes in the access
control mechanisms, not new ciphers.

You should probably look at some of the standard references on cryptography. At
least read Schneier's "Applied Cryptography", which should be in any good library
if you cannot currently afford your own copy.

One list of online resources is:
http://www.freeswan.org/freeswan_trees/freeswan-1.8/doc/web.html#crypto.link

--
You have received this message because you are subscribed to the selinux list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.