Why I Say Ada is the Ideal Language for Cryptography.

[Austin Obyrne <austin.obyrne@hotmail.com>]

As a very much ‘hands-on’ ships’ engineer (far removed from academia) I immediately spotted the benefit of being able to fragment any size of problem into smaller packages with all the attendant benefits that that implies when I was first introduced to Ada-83 from ‘Basic’ and ‘C’.  It was that marvellous attribute that captivated me to Ada-83.

In my current vector cryptography I use very large arrays of keys that are stored as multiple ‘Case’ Statements containing as many as 15000 elements in a single Ada Package.  Effectively these are what I call ‘off-page’ (reminiscent of flowchart nomenclature) package containers.

When I try writing my ciphers in other languages and look for corresponding equivalent containers its ZILCH and that’s just one aspect of the comparison.  That puts me off for ever more in using other languages – especially interpreter-only languages.

The fact that NSA owns the Ada language albeit GPL is a door I won’t have to force open either.

An array of vector coefficients that contains 15000 non-repeating elements can be shuffled in 15000! (15000 factorial) permutations of order.  In my crypto scheme there are 3 such packages for each vector that can be further shuffled in 3 factorial ways.  There are two vector based ‘Keys” and two scalar based Keys i.e. 8 packages in my algorithm – the elements of each package are called sequentially for the encryption of each and every fresh plaintext. 

A different combination of shuffling keys (trans-positional moves) and slice keys (start of calling sequence) may be used for every encryption session without reuse in billions of sessions and that is only the beginning of the available entanglement.  The upshot is a ciphertext string that is 99.4% non repeating, leading to another key in itself. 

It goes without saying almost that the combination of Vectors and Ada Packages is the ideal environment for encryption.

In passing, the vector ‘cross product’ in Ada computations used to have to be done longhand using the determinant and its minors when Ada-83 first came out but there was such a 'hue and cry' for a function to do it automatically that it was built into the Ada library in 2012 ? (I think)- anybody?

I prefer and use the older expanded determinant method although I have customised that also to my own liking.

There is a wonderful symbiosis between Ada and Vector methodologies in this cryptography - I feel fortunate to have realised this at the time of my first intorduction to Ada albeit not related to cryptography yet at that time.

Complements of the Season to everybody.

Austin