Re: "Small" problem

From: Michael Amling (nospam_at_nospam.com)
Date: 08/16/03

  • Next message: Benjamin Choi: "New stream cipher algorithm, Adaptive Stream Cipher"
    Date: Sat, 16 Aug 2003 03:32:55 GMT
    
    

    Gregory G Rose wrote:
    > [I'm replying to both Alex's and Mike's posts,
    > having never received Mike's.]
    >
    > In article <W7a%a.732212$ro6.15056543@news2.calgary.shaw.ca>,
    > Alex Flanagan <spiffy43@hotmail.com> wrote:
    >
    >>Mike Amling wrote:
    >>>
    >>> You mean if two states for the same date hashed to the same value,
    >>>and the wrong one is found first by the 0..255 loop?
    >
    > Yes, exactly.
    >
    >>> You can avoid this possibility by choosing a k for which it doesn't
    >>>happen. 2**18 messages have to be checked to make sure than no two
    >>>states for a given date produce the same hash. It should take only a few
    >>>seconds on a PC to vet a given k.
    >
    > Not really; that will be true for any given
    > *date*, but you won't be able to choose such a k
    > that will work for arbitrary dates.

       The plan (This not Luby-Rackov. This is the hash plan.) is to send
    the 26 least significant bits of the HMAC_k of (10-bit date concatenated
    with 8-bit state). Dates are allowed to be arbitrary only to the extent
    that they fit into 10 bits. A key can be vetted by comparing all 2**18
    possible different hashes produced for a given k and rejecting k if
    there are any duplicates in the least significant 26 bits. How do you
    arrive at "Not really"?

    --Mike Amling


  • Next message: Benjamin Choi: "New stream cipher algorithm, Adaptive Stream Cipher"

    Relevant Pages

    • Re: braindump
      ... Dang. ... I will read all the posts in a thread before replying. ... If you have any suitable food, give it to the guard on duty. ...
      (microsoft.public.cert.exam.mcse)
    • John Jones, DOS attacker, Stop jamming Google newsreader with the malicious subject lines
      ... PENTHO NO-MATES VALEY you LITTTLE SNOT STOP REPLYING TO YOUR OWN ... POSTS AND STOP CROSSPOSTING TO SCI LOGIC 1 new of 57 John Jones ...
      (sci.logic)
    • health by means of brown soldier
      ... If does Hussein accelerate so slightly, whenever Rose assumes the crazy breed very least? ... It might loosely lift competent and tackles our big, necessary healths as yet a zone. ... Gregory, still formulating, swells almost inside, as the equivalent nominates near their pub. ...
      (sci.crypt)
    • Re: Photos
      ... session with Trudi.. ... If some are choosing the option of only seeing unread messages, ... >>> Please QUOTE the post you are replying to. ... >>> the thread without necessarily seeing all the posts. ...
      (microsoft.public.windowsxp.newusers)
    • Re: JSH: Measuring post impact
      ... I indeed find JSH's blog & some archived posts well-represented on the first ... Google's "Cached" link for one of the JSH hits gives dramatic colorized ... trying to push one's rankings up, ... replying, if there is no research interest for me. ...
      (sci.math)