A matter of interest.

A matter of interest.

[Austin Obyrne]

I have been experimenting in the last few days with a new version of some vector cryptography that I invented some years ago.

The particular cipher computer program I am referring to is written in Ada-95 for MS Windows XP Vista 7 & 8 (*runs on hand held tablets MS Surface Pro 2 & 3)

My normal test programs that I use as test/tutorials are named PlainTextFile_1.dat increasing in sizes 9,100, 500,1000,2000,4000,10000, 30000 where the size of the file is that number of ASCII plaintext characters for encryption in that test file.

For good measure I decided to add another file of 50000 characters in this new version under test but where to find a ready-made suitable test file of that size was my problem - keying in a file from scratch was just too time - consuming.  I looked around and found I had a file of Ada-95 source code in my  computer and decided to use that.

I want to report what happened next in case it might be useful to programmers of Ada.

The file of 49471 Ada sourcecode characters encrypted and decrypted perfectly which I knew it would anyway so I decided next I would test the fidelity (a privately coined word that I think is apt but is not normal crypto jargon just for your info) of the entire experiment in cryptology by test-running the decrypted source code as (message-text now) in my Ada-Gide editor - it compiled and ran perfectly without any editing and immediately encrypted some test programs itself.  

I am very pleased with this result for obvious reasons - compiling first time round in an Ada compiler must be the most stringent test of unadulterated transport imaginable.

The reason I am posting this article is to make these points :-

*1) Programmers of any language may remotely post their stuff verbatim to others with impunity using Ada as the carrier.

*2) Cryptographers may use Ada to send encrypted ciphers to others verbatim and know that they will work - there is no more secure method of secure transport and they will only need to periodically update the keys (which may be public or encrypted as preferred, periodically) the initial delivery of the mutual database must be made in a one-off secure courier delivery of course.

I hope it is in order for me to post a crypto article here bearing in mind that you guys are in a programming language usenet.

The host cipher is on offer to all.

Regards to all,

Adacrypt.

Re: A matter of interest.

[Simon Wright]

Austin Obyrne <austin.obyrne@hotmail.com> writes:

> The file of 49471 Ada sourcecode characters encrypted and decrypted
> perfectly which I knew it would anyway so I decided next I would test
> the fidelity (a privately coined word that I think is apt but is not
> normal crypto jargon just for your info) of the entire experiment in
> cryptology by test-running the decrypted source code as (message-text
> now) in my Ada-Gide editor - it compiled and ran perfectly without any
> editing and immediately encrypted some test programs itself.

This is good, but as a method of testing I'd have used a byte-by-byte
comparison of original/decrypted files. In a Unix environment, I'd use
cmp(1).

http://unixhelp.ed.ac.uk/CGI/man-cgi?cmp

Re: A matter of interest.

[Austin Obyrne]

On Sunday, November 30, 2014 12:49:41 PM UTC, Simon Wright wrote:
> Austin Obyrne <austin.obyrne@hotmail.com> writes:
> 
> > The file of 49471 Ada sourcecode characters encrypted and decrypted
> > perfectly which I knew it would anyway so I decided next I would test
> > the fidelity (a privately coined word that I think is apt but is not
> > normal crypto jargon just for your info) of the entire experiment in
> > cryptology by test-running the decrypted source code as (message-text
> > now) in my Ada-Gide editor - it compiled and ran perfectly without any
> > editing and immediately encrypted some test programs itself.
> 
> This is good, but as a method of testing I'd have used a byte-by-byte
> comparison of original/decrypted files. In a Unix environment, I'd use
> cmp(1).
> 
> http://unixhelp.ed.ac.uk/CGI/man-cgi?cmp

But surely the compiler does a bit-by-bit job ?

For the purpose either of these is excellent !
Thanks for your info - Austin.

Re: A matter of interest.

[Björn Lundin]

On 2014-11-30 14:24, Austin Obyrne wrote:

> 
> But surely the compiler does a bit-by-bit job ?

no

 Year : Integer := 2014;

will compile just as

 YEAR : INTEGER := 2014;

and

 YEAR           : INTEGER               := 2014;

but they are of course different when represented
as files.


--
Björn

Re: A matter of interest.

[brbarkstrom]

On Sunday, November 30, 2014 11:08:06 AM UTC-5, björn lundin wrote:
> On 2014-11-30 14:24, Austin Obyrne wrote:
> 
> > 
> > But surely the compiler does a bit-by-bit job ?
> 
> no
> 
>  Year : Integer := 2014;
> 
> will compile just as
> 
>  YEAR : INTEGER := 2014;
> 
> and
> 
>  YEAR           : INTEGER               := 2014;
> 
> but they are of course different when represented
> as files.
> 
> 
> --
> Björn

I do not believe that a new cryptographic algorithm can be trusted
until it has been published and attacked by competent cryptanalysts.
The classic text is

Schneier, B., 1996: Applied Cryptography: Protocols, Algorithms,
and Source Code in C, John Wiley & Sons, New York, NY

If the algorithm hasn't been published, the usual term would be
"security by obscurity."  It doesn't work.

Bruce B.

Re: A matter of interest.

[Simon Wright]

brbarkstrom@gmail.com writes:

> I do not believe that a new cryptographic algorithm can be trusted
> until it has been published and attacked by competent cryptanalysts.
> The classic text is
>
> Schneier, B., 1996: Applied Cryptography: Protocols, Algorithms,
> and Source Code in C, John Wiley & Sons, New York, NY
>
> If the algorithm hasn't been published, the usual term would be
> "security by obscurity."  It doesn't work.

To be fair to Austin, he has published his work-in-progress.

Re: A matter of interest.

[brbarkstrom]

On Sunday, November 30, 2014 1:12:47 PM UTC-5, Simon Wright wrote:
> brbarkstrom writes:
> 
> > I do not believe that a new cryptographic algorithm can be trusted
> > until it has been published and attacked by competent cryptanalysts.
> > The classic text is
> >
> > Schneier, B., 1996: Applied Cryptography: Protocols, Algorithms,
> > and Source Code in C, John Wiley & Sons, New York, NY
> >
> > If the algorithm hasn't been published, the usual term would be
> > "security by obscurity."  It doesn't work.
> 
> To be fair to Austin, he has published his work-in-progress.

Fair enough.  However, I think he should use the available Ada versions
of standard encryption algorithms.  Those will save his time and worries
about security by the user community.  While I have a PhD in astronomy and 
had several years of advanced math as an undergraduate, I do not have the 
mathematical qualifications to build a new encryption algorithm.  Folks
who do that kind of work need to have a very deep understanding of number
theory and related subjects.  'Nuff said.

Bruce B.

Re: A matter of interest.

[Denis McMahon]

On Sun, 30 Nov 2014 11:56:09 -0800, brbarkstrom wrote:

> Fair enough.  However, I think he should use the available Ada versions
> of standard encryption algorithms.  Those will save his time and worries
> about security by the user community.  While I have a PhD in astronomy
> and had several years of advanced math as an undergraduate, I do not
> have the mathematical qualifications to build a new encryption
> algorithm.  Folks who do that kind of work need to have a very deep
> understanding of number theory and related subjects.  'Nuff said.

Generic +1

Dan Boneh at Stanford runs online cryptography courses. The biggest thing 
I learned from his courses was "leave it to the experts, because well 
meaning amateurs haven't a hope in hell of getting it right"

-- 
Denis McMahon, denismfmcmahon@gmail.com

Re: A matter of interest.

[Austin Obyrne]

On Monday, December 1, 2014 4:25:17 PM UTC, Denis McMahon wrote:
> On Sun, 30 Nov 2014 11:56:09 -0800, brbarkstrom wrote:
> 
> > Fair enough.  However, I think he should use the available Ada versions
> > of standard encryption algorithms.  Those will save his time and worries
> > about security by the user community.  While I have a PhD in astronomy
> > and had several years of advanced math as an undergraduate, I do not
> > have the mathematical qualifications to build a new encryption
> > algorithm.  Folks who do that kind of work need to have a very deep
> > understanding of number theory and related subjects.  'Nuff said.
> 
> Generic +1
> 
> Dan Boneh at Stanford runs online cryptography courses. The biggest thing 
> I learned from his courses was "leave it to the experts, because well 
> meaning amateurs haven't a hope in hell of getting it right"
> 
> -- 
> Denis McMahon, denismfmcmahon@gmail.com

Hi Dennis,

It is hard to comment on your remarks without seeming to have an axe to grind. There is nothing in it for me to sound off but I must say for whatever you think its worth - the monumental mistake has been in leaving to these self acclaimed experts - they have had fifty years of trying since the inception of number-theoretic computer driven cryptography in the seventies and so far nothing.

They are so far down the wrong road they do not realise  it.

The nub of the matter is that our beautiful number system that is so good at what it is intended to do is totally unsuited to cryptography because of the sheer transparency of it that is impossible to hide by any normal linear means. Nobody seems to realise this and that alternatives must be found - going down the road of intractability through complexity is both futile and by the way non-mathematical

The most fertile areas for research are far away from the traditional number system but that is as far as those people can think - hands on problem solving like finding crypto algorithms is beyond their ken it appears.

They are even talking about "principles of cryptography" which cannot exist until proper cryptography i.e. "Theoretically Unbreakable" algorithms are demonstrated  They are also talking about the mathematics of cryptography which does not exist for the same reason.

I can demonstrate my invention of this ultimate class of cryptography and have made it available on my website http://www.adacryptpages.com for all to see and criticise.  The establishment are ignoring all overtures from me to submit my stuff and only one of about six has had the decency to answer my email - That is the "Imperial College  London" - Professor Craster is Head of the Applied Maths department there.

The establishment are a dishonest, intransigent lot of bluffers.

I think we may be on the cusp of a massive change so scared is everybody to come out and talk.

It is ludicrous to see the list of papers, past, present and future on complexity-theoretic algorithms that have no future being posted on the websites of many universities and even the IACR (International association for Cryptological Research) - I think.  

No Sir - I do not think much of your suggestions that we leave it to these people.

adacrypt

Re: A matter of interest.

[Jeffrey Carter]

My experience with Obyrne is that he asks for help but does not accept it.

He only knows how to read text files, so he can only encrypt text files. In a
previous thread he was shown how to read, write, and manipulate bytes from any
kind of file, so he concluded that he could only encrypt text files.

In this thread he refused to provide enough information for anyone to help him,
and then concluded that he couldn't do whatever it is he was trying to do.

I am not qualified to comment on his cryptographic ability, but I have noticed a
total absence of news stories on a new cryptographic technique that will
revolutionize the field.

-- 
Jeff Carter
"Apple juice and Darvon is fantastic together."
Play It Again, Sam
127

Re: A matter of interest.

[David Botton]

Austin,

Have you considered creating a contest to break your encryption? Enough prize money and you can get an entire convention of hackers to validate your ideas.

For example:
https://telegram.org/blog/cryptocontest

David Botton

Re: A matter of interest.

[mrvmurray]

On Tuesday, 2 December 2014 02:24:54 UTC, David Botton  wrote:
> Austin,
> 
> Have you considered creating a contest to break your encryption?
> Enough prize money and you can get an entire convention of hackers
> to validate your ideas.

He did , and lost. In typical form, he did not recognise this.

The cipher as stated is brute-forceable in seconds, but the contest as
conducted by O'Byrne was flawed. In short, he cheated, and created
a ciphergram that his "Bob" could not read using the agreed settings.

However, his cipher creates output that heavily leaks the parameters that
he was changing, so obtaining these are not hard work, just an irritant.

M

Re: A matter of interest.

[Austin Obyrne]

On Monday, December 1, 2014 11:49:15 PM UTC, Jeffrey Carter wrote:
> My experience with Obyrne is that he asks for help but does not accept it.
> 
> He only knows how to read text files, so he can only encrypt text files. In a
> previous thread he was shown how to read, write, and manipulate bytes from any
> kind of file, so he concluded that he could only encrypt text files.
> 
> In this thread he refused to provide enough information for anyone to help him,
> and then concluded that he couldn't do whatever it is he was trying to do.
> 
> I am not qualified to comment on his cryptographic ability, but I have noticed a
> total absence of news stories on a new cryptographic technique that will
> revolutionize the field.
> 
> -- 
> Jeff Carter
> "Apple juice and Darvon is fantastic together."
> Play It Again, Sam
> 127

Hi Jeffrey,

I'm going to be very brief because we are on someone else's usenet.

My modus operandi is totally focused on getting the core algorithm right - that must be the beginning middle and end of cryptography in my view.  Reading in data for encryption is easy and if it's not then there are millions of people out there (some right here in this group) who are expert in helping out with their expertise that I can call on albeit they know little of cryptography.  

I have only this very weekend completed a profoundly secure algorithm that I would promote to Gauss, Archimedes or Euclid with impunity if they were still here.

I can call on millions of people out there when I need them for the peripheral task of say encrypting images, or music - and its much more than just the reading in bit usually to do this since you imply that that is all there is to it.

Witness the posts in sci crypt - a plethora of management incidents - everything you see posted is to do with difficult infra structure problems that emanate from a *weak core intelligence.

I don't read in as one might do using AES say - no , I am playing to different goalposts - there is method in my seeming madness and it is nothing to do with asking and then refusing help from readers - I shamelessly pick peoples' brains at times just for inspiration.

Meanwhile its good to hear - Austin

Re: A matter of interest.

[Dennis Lee Bieber]

On Mon, 1 Dec 2014 18:43:00 -0800 (PST), Austin Obyrne
<austin.obyrne@hotmail.com> declaimed the following:

>I have only this very weekend completed a profoundly secure algorithm that I would promote to Gauss, Archimedes or Euclid with impunity if they were still here.
>
	You've rediscovered the "one time pad" or Caesar cipher?

	That's about the level of encryption that would have been amazing to
the three so named.

	To them, the now weak DES would have required weeks of work, even with
the key and a description of how the algorithm works, just to encode a
message.
-- 
	Wulfraed                 Dennis Lee Bieber         AF6VN
    wlfraed@ix.netcom.com    HTTP://wlfraed.home.netcom.com/

Re: A matter of interest.

[Eryndlia Mavourneen]

On Sunday, November 30, 2014 1:56:10 PM UTC-6, brbar...@gmail.com wrote:
> On Sunday, November 30, 2014 1:12:47 PM UTC-5, Simon Wright wrote:
> > brbarkstrom writes:
> > 
> > > I do not believe that a new cryptographic algorithm can be trusted
> > > until it has been published and attacked by competent cryptanalysts.
> > > The classic text is
> > >
> > > Schneier, B., 1996: Applied Cryptography: Protocols, Algorithms,
> > > and Source Code in C, John Wiley & Sons, New York, NY
> > >
> > > If the algorithm hasn't been published, the usual term would be
> > > "security by obscurity."  It doesn't work.
> > 
> > To be fair to Austin, he has published his work-in-progress.
> 
> Fair enough.  However, I think he should use the available Ada versions
> of standard encryption algorithms.  Those will save his time and worries
> about security by the user community.  While I have a PhD in astronomy and 
> had several years of advanced math as an undergraduate, I do not have the 
> mathematical qualifications to build a new encryption algorithm.  Folks
> who do that kind of work need to have a very deep understanding of number
> theory and related subjects.  'Nuff said.
> 
> Bruce B.

As I recall, the situation is reversed:  Austin is a mathematician and not a programmer much less a professional Ada developer.

Re: A matter of interest.

[mrvmurray]

On Monday, 1 December 2014 19:36:39 UTC, Eryndlia Mavourneen  wrote:
> As I recall, the situation is reversed:  Austin is a mathematician and not a programmer much less a professional Ada developer.

He's not a mathematician. He claims to be a retired marine engineer.

M