The Changing Face of Future Cryptography.

[Austin Obyrne]

Future cryptography may well be all about finding transitory substitutions for the traditional data sets and using the appropriate methodology of the temporary substitution as the encryption means in cryptography.

I have used ‘displacement’ as one substitution but there may be lots more physical analogues out there waiting for the finding by well-motivated crypto researchers.

This change is a watershed that simply must come – there is too much effort being spent at present on masking the natural transparency of the traditional number line (without complete success) as the selection domain for use as the raw data in encryption transformations being made by design cryptographers.

I think it is safe to say that all contemporary ciphers will be sent to the scrap heap sooner than we think for this reason.

I repeat, cryptography is an industry that goes out – not up – there are no principles to be salvaged from defunct cryptography that would suggest retaining them as building blocks for future ciphers that might say cryptography goes up - it doesn't.

Haranguing people for a lack of interest and knowledge of current old ciphers is small-minded and parochial – old ciphers are even more useless than old newspapers – the latter can be recycled at least.

A lot of noses in the establishment will be put out of joint in the changeover to this new approach but that’s just too bad – it has to happen.

- adacrypt

[xxein]

xxein: We have discussed this before. It all boils down to a
character set. There is no reason to elevate and bring it down
again. The manipulations all lead to the same place, text back to
text.

Here's a example: dd]@0&&zkSJ/zzuf]QQ7xoXB8}q
ee^A1''{llTK0{{vg^RR8ypYC9 r

The only difference is that the first text encoded is missing a space
between two words. Let's try padding.

93-VVO2"vvl]]E<!llgXOCC)jaJ4*oc A sinple letter repeat at the
beginning.

Let's try this again. I have done it before so I do not fear
government crypto interception analysis.

3{4d;W&u;7`6HT\IRF`}z],1IAK1y%R*=MurNXd6Nf"#Dn,W:%]k!`3gH1sE9%3t2%Wt(b;
8g[Qt.ae?Jj%Jrw"9 I do not fear any decoding. But if the government
did decode, they would realize that this is only a test.

Now it is -*>$?v|s0G)?g=yDv> }yXdu@a.?:?q{:g^R"Z@V&lnq7'3j&
+hUV8K4\mb)BGZPbg@6>"=U>yLBP3x9cOT rkm(

Any padding could certainly obscure this. But as I have already told
you this before, this en-decryption program can be modified in a
surreptitious manner. By missing any message, it is unreadable in
plus fashion.

Yet (if I remember how I wrote it some 20 yrs. ago ) it uses only 92
characters in the ASCII numerical. I beg of anybody to decipher what
I wrote.

[Austin Obyrne]

On Thursday, May 23, 2013 2:14:36 AM UTC+1, xxein wrote:
> On May 22, 5:11 am, Austin Obyrne <austin.oby...@hotmail.com> wrote: > Future cryptography may well be all about finding transitory substitutions for the traditional data sets and using the appropriate methodology of the temporary substitution as the encryption means in cryptography. > > I have used ‘displacement’ as one substitution but there may be lots more physical analogues out there waiting for the finding by well-motivated crypto researchers. > > This change is a watershed that simply must come – there is too much effort being spent at present on masking the natural transparency of the traditional number line (without complete success) as the selection domain for use as the  raw data in encryption transformations being made by design cryptographers. > > I think it is safe to say that all contemporary ciphers will be sent to the scrap heap sooner than we think for this reason. > > I repeat, cryptography is an industry that goes out – not up – there are no principles to be salvaged from defunct cryptography that would suggest retaining them as building blocks for future ciphers that might say cryptography goes up - it doesn't. > > Haranguing people for a lack of interest and knowledge of current old ciphers is small-minded and parochial – old ciphers are even more useless than old newspapers – the latter can be recycled at least. > > A lot of noses in the establishment will be put out of joint in the changeover to this new approach but that’s just too bad – it has to happen. > > - adacrypt xxein: We have discussed this before. It all boils down to a character set. There is no reason to elevate and bring it down again. The manipulations all lead to the same place, text back to text. Here's a example: dd]@0&&zkSJ/zzuf]QQ7xoXB8}q ee^A1''{llTK0{{vg^RR8ypYC9 r The only difference is that the first text encoded is missing a space between two words. Let's try padding. 93-VVO2"vvl]]E<!llgXOCC)jaJ4*oc A sinple letter repeat at the beginning. Let's try this again. I have done it before so I do not fear government crypto interception analysis. 3{4d;W&u;7`6HT\IRF`}z],1IAK1y%R*=MurNXd6Nf"#Dn,W:%]k!`3gH1sE9%3t2%Wt(b; 8g[Qt.ae?Jj%Jrw"9 I do not fear any decoding. But if the government did decode, they would realize that this is only a test. Now it is -*>$?v|s0G)?g=yDv> }yXdu@a.?:?q{:g^R"Z@V&lnq7'3j& +hUV8K4\mb)BGZPbg@6>"=U>yLBP3x9cOT rkm( Any padding could certainly obscure this. But as I have already told you this before, this en-decryption program can be modified in a surreptitious manner. By missing any message, it is unreadable in plus fashion. Yet (if I remember how I wrote it some 20 yrs. ago ) it uses only 92 characters in the ASCII numerical. I beg of anybody to decipher what I wrote.

I think you are basically agreeing in broad priciple with what I am saying. The domain for selection of the raw data for encryption transformations needs to be something less transparent than the current use of consecutive numbers, alpahabetics and well ordered alphanumerics that is mutually known to both cryptographer and cryptanalyst.

<There is no reason to elevate and bring it down

<again. The manipulations.....

I doubt that this is generally true however and depends on the methods being used by individuals - but this is just a lesser aside in any case - it is not highly important.

Clearly, there is a need for *some change.

The present system in which the cryptanalyst knows the selection domain being used by the cryptographer is loaded against the cryptogrpaher from the start and makes his job impossible.

Current research is about finding complexity-theoretic algorithms that will tip the balance back in favour of the cryptographer but I contend that is just too difficult to enact - it's a lost cause - it is unnecessary anyway as there are better solutions, to wit your own demonstration here and the one I am promoting.

Hope I am reading you correctly.

Best Wishes - Austin O' Byrne

[Austin Obyrne]

On Thursday, May 23, 2013 2:14:36 AM UTC+1, xxein wrote:

> On May 22, 5:11 am, Austin Obyrne <austin.oby...@hotmail.com> wrote: > Future cryptography may well be all about finding transitory substitutions for the traditional data sets and using the appropriate methodology of the temporary substitution as the encryption means in cryptography. > > I have used ‘displacement’ as one substitution but there may be lots more physical analogues out there waiting for the finding by well-motivated crypto researchers. > > This change is a watershed that simply must come – there is too much effort being spent at present on masking the natural transparency of the traditional number line (without complete success) as the selection domain for use as the  raw data in encryption transformations being made by design cryptographers. > > I think it is safe to say that all contemporary ciphers will be sent to the scrap heap sooner than we think for this reason. > > I repeat, cryptography is an industry that goes out – not up – there are no principles to be salvaged from defunct cryptography that would suggest retaining them as building blocks for future ciphers that might say cryptography goes up - it doesn't. > > Haranguing people for a lack of interest and knowledge of current old ciphers is small-minded and parochial – old ciphers are even more useless than old newspapers – the latter can be recycled at least. > > A lot of noses in the establishment will be put out of joint in the changeover to this new approach but that’s just too bad – it has to happen. > > - adacrypt xxein: We have discussed this before. It all boils down to a character set. There is no reason to elevate and bring it down again. The manipulations all lead to the same place, text back to text. Here's a example: dd]@0&&zkSJ/zzuf]QQ7xoXB8}q ee^A1''{llTK0{{vg^RR8ypYC9 r The only difference is that the first text encoded is missing a space between two words. Let's try padding. 93-VVO2"vvl]]E<!llgXOCC)jaJ4*oc A sinple letter repeat at the beginning. Let's try this again. I have done it before so I do not fear government crypto interception analysis. 3{4d;W&u;7`6HT\IRF`}z],1IAK1y%R*=MurNXd6Nf"#Dn,W:%]k!`3gH1sE9%3t2%Wt(b; 8g[Qt.ae?Jj%Jrw"9 I do not fear any decoding. But if the government did decode, they would realize that this is only a test. Now it is -*>$?v|s0G)?g=yDv> }yXdu@a.?:?q{:g^R"Z@V&lnq7'3j& +hUV8K4\mb)BGZPbg@6>"=U>yLBP3x9cOT rkm( Any padding could certainly obscure this. But as I have already told you this before, this en-decryption program can be modified in a surreptitious manner. By missing any message, it is unreadable in plus fashion. Yet (if I remember how I wrote it some 20 yrs. ago ) it uses only 92 characters in the ASCII numerical. I beg of anybody to decipher what I wrote.

Just a thought,

I’m taking it that this is just a demonstration using an incomplete example of a scrambled selection domain that demonstrates what you have in mind.

In real life you would probably opt for a larger scrambled (but reversible) version of ASCII ?

Considering the three basic attacks by a cryptanalyst I think you might get away with,

1) direct attack on the ciphertext i.e. by inversion.

Statistical attacks by,

2) analysis of the frequency distribution in a Babbage-Kasiski attack and

3) statistical mapping (much the same thing – your selection domain of alphanumerics is almost bound to become known) emanating from a known plaintext sample.

Cheers.

- adacrypt

[Gordon Burditt]

> This change is a watershed that simply must come – there is too
> much effort being spent at present on masking the natural transparency
> of the traditional number line (without complete success) as the
> selection domain for use as the raw data in encryption transformations
> being made by design cryptographers.

What you seem to be saying here is that if the *PLAINTEXT* is in
ASCII and is English text, then *GAME OVER*, you (or anyone else)
is going to fail trying to keep it secret by encrypting it. That
seems to be throwing your ciphers under the bus along with everyone
else's. It also seems to mean that no matter how much one pre-scrambles
that plaintext before getting on to the main encryption (like, say,
translating it to Egyptian hieroglyphics), you're still going to
fail to keep it secret.

You don't get to control plaintext.

> I repeat, cryptography is an industry that goes out – not up –
> there are no principles to be salvaged from defunct cryptography
> that would suggest retaining them as building blocks for future
> ciphers that might say cryptography goes up - it doesn't.

There are a couple of things you might learn from old ciphers.
A few of these are definitions of terms you don't seem to
know the meaning of but use anyway:
plaintext
ciphertext
key (a message is encrypted using one key, although you
may split that into as many subkeys as you like).

Every time someone tries to figure out about how many bits are in
the key you are using, you dodge the question. At one point you
said that you could set up a shared key by exchanging two integers
(presumably 16-bit or less, from your examples).

[Austin Obyrne]

On Thursday, May 23, 2013 9:18:37 AM UTC+1, Gordon Burditt wrote:
> > This change is a watershed that simply must come – there is too > much effort being spent at present on masking the natural transparency > of the traditional number line (without complete success) as the > selection domain for use as the raw data in encryption transformations > being made by design cryptographers. What you seem to be saying here is that if the *PLAINTEXT* is in ASCII and is English text, then *GAME OVER*, you (or anyone else) is going to fail trying to keep it secret by encrypting it. That seems to be throwing your ciphers under the bus along with everyone else's. It also seems to mean that no matter how much one pre-scrambles that plaintext before getting on to the main encryption (like, say, translating it to Egyptian hieroglyphics), you're still going to fail to keep it secret. You don't get to control plaintext. > I repeat, cryptography is an industry that goes out – not up – > there are no principles to be salvaged from defunct cryptography > that would suggest retaining them as building blocks for future > ciphers that might say cryptography goes up - it doesn't. There are a couple of things you might learn from old ciphers. A few of these are definitions of terms you don't seem to know the meaning of but use anyway: plaintext ciphertext key (a message is encrypted using one key, although you may split that into as many subkeys as you like). Every time someone tries to figure out about how many bits are in the key you are using, you dodge the question. At one point you said that you could set up a shared key by exchanging two integers (presumably 16-bit or less, from your examples).

Your'e talking about several different things here - you're mixing scalars with vectors , binary with denary , in simultaneous applications to two different ciphers.

As I said already in another reply its is impossible for me to comment on isolated fragments of something that requires a lot of one-to-one with a lot of chalk 'n talk thrown in.

I can assure you this crypto is thoroughly coherent through and through and is proven all the way.

Also, who cares about pedantic little herd rules that you insist must be obeyed - I don't - a succesful cipher doesn't need them and may even make its own rules for others to follow.

You cannot break into this cryptography at some elevated level and think you don't need what goes before on the back of knowing some old scalar crypto from the past - they don't gel - they are as chalk is to cheese.

I could be here for the next ten years correcting your wrong assumptions.

Fair question : Are you totally au fait with vector methods and plane geometry or not?

You cannot go forward without this - it is disrespectful to try - why don't you ask some one you trust for an honest opinion as to whether you are equipped for it or not.

Hacking away with no tools and thinking you can bluff and gatecrash your way into solidly established algorithms is being dishonest - to yourself - nobody else.

- adacrypt

[SG]

On May 23, 10:18 am, Gordon Burditt wrote:
> What you seem to be saying here is that if the *PLAINTEXT* is in
> ASCII and is English text, then *GAME OVER*, you (or anyone else)

As far as I understand, he suggests that one has to map each plain
text character to an element in a higher dimensional space because he
believes the "natural number line" is somehow "transparent" (whatever
that means). What's interesting about this is that he actually maps
them to a LINE in 3D space where the line's direction vector has only
nonzero coefficients. So, he just adds lots of redundancy. You can
throw away any two of the three dimensions without loss of
information. This would then be equivalent to a "scalar cipher" which
obviously suffers from the same "transparency issues" he believes to
have fixed by doing the vector dance. So, he actually did not improve
anything -- not even by his own standards. I find this hilarious.

You are aware of the fact that he's been doing his "research" for
about 20(*) years by now, right?

(* according to his own words)

[SG]

On May 23, 1:39 pm, Austin Obyrne wrote:
> Fair question : Are you totally au fait with vector methods and plane geometry or not?

Feel free to assume that your opposers know more about math than you
do. The problem with your texts is your way of describing things. You
make a lot of misuse of already existing terms and use made-up terms
you havn't defined properly/at all. Your opposers are not stupid. But
they are definitely not mind readers. It's your job to write something
coherent, something that makes sense to people. Don't you think that
it's more likely that you -- some guy who admitted lack of math/crypto
education -- don't know how to express your ideas than everybody
beeing too stupid to follow your supposedly good ideas? Do something
about your cognitive biases. You are really imagining things. Seek
help.