http://www.perlmonks.org?node_id=97246


in reply to CipherTextI

Here are my objections to this post.
  1. This isn't valid Perl. It won't run.
  2. Even if it was, you are failing to achieve the basic level of encryption taken for granted decades ago. This is an improvement on the xor nonsense that you had previously. But it isn't much of an improvement.
  3. The fact that you are trying to get a patent is but another reason to avoid this. It is possible that you might get a patent. But that is IMO due to how broken the patent system is, and not to the merits of your case.
  4. A basic rule in cryptography is to never trust the crypanalysis of the author of the code. This applies when the author is widely recognized as a competent researcher. It applies doubly when the author is an amateur.
  5. There are widely available free alternatives which are much more trustworthy. For instance RSA is no longer encumbered by a patent, has been analyzed intensely, and can be used for the application that you mention by the simple expedient of putting your website on an https server. That takes care of the details on the server's end, and the browser will do the rest.
  6. You still have not realized that your prototypes are completely being ignored? We went through this before and I know you didn't believe us then, but it is true. You would be better off just removing them all because they are not used on method calls.
In short, I am actively recommending that people not try to use this.

Replies are listed 'Best First'.
Re: Re (tilly) 1: CipherTextI
by merlyn (Sage) on Jul 17, 2001 at 21:19 UTC
    I'm about 3/4ths of the way through The Code Book, and I'm even more amazed at how interesting the task of cryptohacking is, and how much easier it is for the experts to break things than I had ever imagined. Sure, some of it is brilliant strokes, and some of it is dumb luck, but the account of snapping the Enigma machines is a definite page-turner.

    The CiperTextI looks like the stuff these guys solve as entrance exams in first year crypto. {grin}

    -- Randal L. Schwartz, Perl hacker

      For all interested people I thought I'd add The Codebreakers by David Kahn, the first 'popular-scientific' work in that field (originally from 1967 iirc). I don't know how much overlap there is between this book and The Code Book mentioned by merlyn. But it's a good read - that's what I meant by popular-scientific ;)

      -- Hofmator

      To get the facts, the Code book was a facinating read (particularly for its nice links back to historic events).

      To wet the tastebuds (and pass the hours quite nicely) I've always enjoyed Cryptonomicon by N.S. Its link to reality is a little cloudy at times but it certainly fires up the imagination.