Re: [Full-Disclosure] Possibly a stupid question RPC over HTTP

From: Kyle Maxwell (krmaxwell_at_gmail.com)
Date: 10/25/04

  • Next message: Vincent Archer: "Re: [Full-Disclosure] FAKE: RedHat: Buffer Overflow in "ls" and "mkdir""
    To: "Airey, John" <john.airey@rnib.org.uk>
    Date: Sun, 24 Oct 2004 22:30:13 -0500
    
    

    On Fri, 22 Oct 2004 14:50:23 +0100, Airey, John <john.airey@rnib.org.uk> wrote:
    > > -----Original Message-----
    > > From: Kyle Maxwell [mailto:krmaxwell@gmail.com ]
    > > I think you may mean something slightly differently; given any large
    > > prime p, I can factor it completely extremely quickly:
    > >
    > > p = 1 * p
    > >
    > > There are no other factors; this *is* the prime factorization. :) Bill
    >
    > Oh no, the whole security of computing has just fallen over, since you've shown that primes don't exist. What next, proving that black is white and getting run over on a zebra crossing?
    >
    > A prime is defined as being divisible by itself and 1 only, so for the purpose of the definition, 1 is not a factor.

    <flame>
    I was trying to give you the benefit of the doubt in my explanation,
    but your response makes it clear that you're not thinking straight. By
    your (almost correct) definition of prime, the factorization is
    trivial! And yes, 1 is a factor. If you can break the prime into ANY
    other factors, then it's NOT a prime.

    You're talking about solving a problem that DOESN'T EXIST BY
    DEFINITION. Re-read my response -- this time without being stupid --
    and you'll see that I was trying to explain to you that the problem is
    the general factoring of large numbers (into primes for what should be
    obvious reasons). This is NOT the same as factoring large primes as
    that's a solved problem. If this is still difficult to understand, any
    handy grade-school maths book should provide additional explanation.
    Testing for primality, which is a related but different problem, is
    solved, but proving that a number is composite is unfortunately not
    the same as knowing its factors.
    </flame>

    As to the question of whether this is a solved problem: we may have to
    agree to disagree; if it were the NSA, given their past interactions
    with the crypto community, I think it likely that they'd have over
    time moved to another type of cryptography. BTW, brute forcing a key
    does not break the system -- and as others have shown in this thread,
    it's impossible to precompute all the keys unless you've broken every
    single PRNG out there, and that's even less likely.

    -- 
    Kyle Maxwell
    [krmaxwell@gmail.com]
    _______________________________________________
    Full-Disclosure - We believe in it.
    Charter: http://lists.netsys.com/full-disclosure-charter.html
    

  • Next message: Vincent Archer: "Re: [Full-Disclosure] FAKE: RedHat: Buffer Overflow in "ls" and "mkdir""

    Relevant Pages

    • Re: -- Limit of ratio of consecutive primes = 1 ?
      ... Twin primes quickly push the limit to 1 exactly. ... show that I was capable of learning. ... Asking limit P/ Pis something like ... The first limit exists, but proving that it ...
      (sci.math)
    • Re: Gs conjecture and Bs postulate
      ... primes would prove that Goldbach's conjecture is decidable without ... proving whether it's true or false. ... So the "odd Goldbach conjecture" ... After some steps the recursive pattern became too ...
      (sci.math)
    • Re: Gs conjecture and Bs postulate
      ... A proof that every even number greater than 10^1000 is the sum of two ... primes would prove that Goldbach's conjecture is decidable without ... proving whether it's true or false. ... So the "odd Goldbach conjecture" ...
      (sci.math)
    • Re: Gs conjecture and Bs postulate
      ... two primes would prove that Goldbach's conjecture is decidable ... without proving whether it's true or false. ... Or for an even worse example, Skewes' number as ...
      (sci.math)
    • Re: I was right, surrogate factoring proof
      ... primes divide n -- at most I'll have to look through sqrt/ ... But don't forget that's included in the cost of the algorithm! ... if surrogate factoring ever becomes ... two values are congruent modulo p, and hopefully not congruent modulo n/p. ...
      (sci.math)