Re: Non-linear combination

From: Mok-Kong Shen (mok-kong.shen_at_t-online.de)
Date: 11/12/03 

Date: Wed, 12 Nov 2003 11:11:52 +0100

Bryan Olson wrote:
>
> Mok-Kong Shen wrote:
> > A question is whether you think that having the quantities
> > bd, ad, bc, ac (in a sense) fully encompassed in a result,
> > in contrast to the case of ordinary modular multiplication
> > where ac gets ignored, is desirable.
>
> A similar question is whether cardboard or styrofoam makes
> better armor.
>
> [...]
> > I don't yet see how the (in my view) relatively
> > complicated scheme of IDEA is a sensible/appropriate
> > basis to compare with my idea which, as I said, is
> > (trivially) simple (or simply trivial).
>
> Limiting one's study to the trivial strikes me as a bad idea in
> most fields, and especially bad in cryptography.

If one wants computationally more expensive non-linear
combinations, then there are lots available. A conceivably
nice one is the f-function of DES. I want to see ways that
are 'very much' cheaper. That's by itself an interesting
problem in my view (though apparently not for you). For
under circumstances one could, because of cheapness,
afford to use many of these to attempt to achieve the
same purpose as the more expensive ones. On the other
hand, trivially simple techniques doesn't 'imply' they are
valueless. Sometimes they don't come to mind simply because
they appear to be too trivial. One of that genre in crypto
that was recently mentioned in the group was the use of one
bit to reduce the volume of information to be transmitted
in connection with ECC, if I remember correctly. (That got
even a, though silly, patent.)

Anyway, from the content of your two posts, I am
absolutely certain that you are totally disinterested
in the issue of very simple/trivial non-linear
combinations (of nature comparable to my proposal) and
consequently I can logically expect that there would
be no further follow-ups from you in this thread. (Or
perhaps even you have already put me on your kill-file,
as I had repeatedly advised you previously.)

[For the general readers only:] My original post was
actually taking up again a theme discussed in a thread
'Remark on multiplication mod 2^n' that I initiated on
15 Apr 2001. There I suggested to employ, if one uses
assembler, the sum of the two registers of the full-length
multiplication result. Further I mentioned that the
operands could additionally be subjected to bit rotations.
There were several follow-ups there that might be also of
some interest in the present connection.

M. K. Shen
--------------------------------
http://home.t-online.de/home/mok-kong.shen

Relevant Pages

  • Re: Problem with files with names like *com.dll
    ... See below for original post + follow-ups. ... I am still experiencing this problem despite trying absolutely everything I ... operating system which is not working correctly and no-one can tell me why. ...
    (microsoft.public.windowsxp.security_admin)
  • Re: recv() problem?
    ... > The original post didn't say that, and I don't think it's true. ... I use a timeout on my recv because in my application there is no way to ... know when the connection will be closed. ... connectivity or device reboot may cause a loss of ...
    (comp.os.linux.networking)
  • Re: change node on ipconfig
    ... Is this a continuation of your original post (Client setup won't connect)? ... > want to set it up with a VPN connection to our sever. ... > MicrosoftWindows DOS ...
    (microsoft.public.windows.server.sbs)
  • Re: More 11.2.6 Update Woes...
    ... I just reread the url in your original post & it looked a little fishy. ... "The connection to the update server was interrupted. ... my inability to connect and download. ... getting to any other web sites or download sites. ...
    (microsoft.public.mac.office.excel)
  • Re: OT - back from surgery!
    ... Somehow, I didn't get the original post, but I'm getting the follow-ups. ... I hope you have a great recovery! ... I had sinus surgery about 9 years ago to repair a deviated septum, remove a bone spur, and correct an enlarged turbinate, and I couldn't belive what a difference it made! ...
    (rec.crafts.beads)
Quantcast