Re: Axcrypt program

ross_at_nospam.com
Date: 07/17/05 

Date: 17 Jul 2005 13:46:23 +0800

Sebastian Gottschalk <seppi@seppig.de> wrote:

>Due to it's partially written in Delphi, the ABI compatibility is bad. You
>can't compile the C++ part without anything else other than Borland C++
>Builder, or it won't work. No, it's not just the calling convention thing.

I suppose it might be worth pointing that out - but remember that an executable
is provided, so the main role of source code is to allow the public assessment
of the algorithm. It really doesn't matter if a given reader doesn't have the
platform to build it, because that's not how you assess an algorithm.

In case this matter has distracted us, may I rephrase my original question:
is the this implementation of AES OK or not? In other words, is it a safe
thing to use for encrypting personal data? I am not talking about national
security here. I would be happy if the implementation was judged to be sound,
so I can focus on the non-cryptographic issues of key storage, etc, where I can
assess the risks for myself.

thanks
ross

Relevant Pages

  • Re: Yet another substitute for Math.random
    ... matter truly random or pseudo-random, given that the tester is truly ... acting on some - possibly very sophisticated - algorithm. ... pseudo-random) then on the long run player B will loose more games ...
    (comp.lang.javascript)
  • Re: What is the accepted technical definition of the word break?
    ... and an algorithm that is too weak because of a successful ... the normal definition of "broken" is ... that an algorithm meets its *claimed* security goals - no matter how weak ... but this is a different matter. ...
    (sci.crypt)
  • Re: SF: New factoring method
    ... At this point I know it's a matter of quadratic residues ... I wasn't sure what JSH knows. ... but factoring /2 still is not an easy task in general. ... I doubt that JSH has the ability to implement his "algorithm". ...
    (sci.math)
  • Re: MODUNI a Good Key Generator
    ... key are useless. ... Don't use it for cryptography. ... It doesn't matter how long the pseudo-random sequence is. ... it doesn't matter how complex you make the algorithm ...
    (sci.crypt)
  • Re: Accuracy of Orion XT IntelliScopes with Object Locator
    ... enough to "dead on" for government work. ... of the algorithm, to first order. ... The reason it might matter is if the ... The PleiadAtlas Home Page at http://www.astronomycorner.net/pleiadatlas/ ...
    (sci.astro.amateur)