Loading ...
Sorry, an error occurred while loading the content.

Re: [PrimeNumbers] Backtrack?

Expand Messages
  • Jud McCranie
    ... What do you mean by backtracking here? Is it that if you try to factor n+1 or n-1 in a proof and get to a large factor, you try to prove it prime in order
    Message 1 of 7 , Nov 8, 2001
    • 0 Attachment
      At 02:43 AM 11/9/2001 +0100, Marcel Martin wrote:

      >Well, since I still don't know whether I will add a backtracking
      >feature or not to the next Primo release, if ever you have something
      >to say about it, please, say it now.

      What do you mean by backtracking here? Is it that if you try to factor n+1
      or n-1 in a proof and get to a large factor, you try to prove it prime in
      order to make the n+1 or n-1 proof of the original number work?


      +--------------------------------------------------------+
      | Jud McCranie |
      | |
      | 137*2^261147+1 is prime! (78,616 digits, 5/2/00) |
      +--------------------------------------------------------+
    • d.broadhurst@open.ac.uk
      Marcel Martin (famously fast worker famously slow decider:-) ... begin{special_pleading} I say this: Pease retain the bactracking of Primo 1.1.0 as an
      Message 2 of 7 , Nov 8, 2001
      • 0 Attachment
        Marcel Martin
        (famously fast worker
        famously slow decider:-)
        wrote:

        > Well, since I still don't know whether I will add a
        > backtracking feature or not to the next Primo release,
        > if ever you have something to say about it, please, say it now.

        \begin{special_pleading}

        I say this: Pease retain the bactracking of Primo 1.1.0
        as an *option*. The likes of Giovanni, Hans, Greg, Nathan,
        Bouk and I are wise enough to experiment as to
        when to use it. It would be a great shame to throw away
        one of your many achievements in implementing it.

        \end{special_pleading}

        David
      • Jud McCranie
        ... OK. Knuth uses the term backtrack in the example in section 4.5.4. +---------------------------------------------------------+ ...
        Message 3 of 7 , Nov 9, 2001
        • 0 Attachment
          At 06:45 AM 11/9/2001 +0100, Marcel Martin wrote:

          >No. In that case I would surely have told of 'recursivity' :-)

          OK. Knuth uses the term "backtrack" in the example in section 4.5.4.


          +---------------------------------------------------------+
          | Jud McCranie |
          | |
          | Programming Achieved with Structure, Clarity, And Logic |
          +---------------------------------------------------------+
        • Andrey Kulsha
          Hello! ... David is right. I d also add that it would be nice to take into account the number of bits gained in previous tests when backtracking. Best wishes,
          Message 4 of 7 , Nov 9, 2001
          • 0 Attachment
            Hello!

            Marcel Martin wrote:

            > Well, since I still don't know whether I will add a backtracking
            > feature or not to the next Primo release, if ever you have something
            > to say about it, please, say it now.

            David Broadhurst answered:

            > I say this: Pease retain the bactracking of Primo 1.1.0
            > as an *option*.

            David is right. I'd also add that it would be nice to take into account the
            number of bits gained in previous tests when backtracking.

            Best wishes,

            Andrey
          • Hans.Rosenthal@t-online.de
            ... I think that *both* backtracking AND more factorization power is most useful when you are tackling numbers beyond a certain (let s say 3k+) size. Hans
            Message 5 of 7 , Dec 1, 2001
            • 0 Attachment
              Marcel Martin wrote:

              > Well, since I still don't know whether I will add a backtracking
              > feature or not to the next Primo release, if ever you have something
              > to say about it, please, say it now.

              I think that *both* backtracking AND more factorization power is most
              useful when you are tackling numbers beyond a certain (let's say 3k+)
              size.

              Hans
            • d.broadhurst@open.ac.uk
              Marcel Martin wrote ... I didn t dare tell you that I use it all the time, including the advanced set up. I thought you might foam at the mouth :-) if I showed
              Message 6 of 7 , Dec 2, 2001
              • 0 Attachment
                Marcel Martin wrote

                > There were very few answers to my post.

                I didn't dare tell you that I use it all the time,
                including the advanced set up.

                I thought you might foam at the mouth :-)
                if I showed you:

                Binary size = 7560
                Running time 53h 38mn 42s for path
                Running time 11h 0mn 50s for rest
                with backtrack at end of Run 1 Part 2
                at 1 GHz

                Thanks for letting me be so "inefficient" :-)

                David
              • d.broadhurst@open.ac.uk
                PS: I looked to see how many backtracks there were in ... answer = 15 backtracks in 353 steps For my money, that was a small price to pay, at this bitsize, for
                Message 7 of 7 , Dec 2, 2001
                • 0 Attachment
                  PS: I looked to see how many backtracks there were in

                  > Binary size = 7560
                  > Running time 53h 38mn 42s for path

                  answer = 15 backtracks in 353 steps

                  For my money, that was a small price to pay,
                  at this bitsize, for complete avoidance of Run 2.
                  Opinions may vary, but experiment should
                  at least be allowed, please, Marcel.

                  David
                Your message has been successfully submitted and would be delivered to recipients shortly.