Re: Surrogate factoring and the k/T ratio

On 24-02-2007 18:20, jstevh@xxxxxxxxx wrote:

I have talked about this idea and approach on the newsgroups before as
I came across it last year, with a few problems still figuring out
details, but the essence has been around for months, and there were
posters claiming it worked no better than random.
I refused to test it myself while I considered theoretical issues and
worried about other things, but finally sat down and checked myself,
and noticed they were wrong, but there is some interesting behavior
around how big k is relative to T--the k/T ratio.
That is, I checked with very small primes--to see if the idea worked
at all--so I used three digit primes, and found that it factored
remarkably well, but I did do other checks and saw it factoring rather
badly, though it still would factor.
Thinking about it, I realized that the k/T ratio was dropping, and
that when it was factoring very well the k/T ratio was about 0.2 or
20%. But I don't know if that is the prime value, as the research is
very rough at this point. But an optimal k/T ratio seems key.
Good for you. Now, what's next? Are you going to factor the number
RSA-704 and win $30,000?


No. The research is very basic at this point.

I've simply verified that the technique is better than random and
appears to give room for optimization.

It'd be nice if you went from a basic idea to best in the world
overnight--like in a Hollywood movie--but the real world doesn't work
that way.

That is the one big surprise and tells me why the newsgroups are not a
good resource or a place to do research--only to announce results,
maybe--as, well, people lie on them.
Did you notice that you are posting this at a newsgroup?


Yup. I'm announcing the result here. But I don't expect any useful
feedback in terms of further development, and I do expect posters to
lie about the research--or raise unreasonable expectations--like you
just tried.

I am fascinated by your accusation. Except for the sentence «Good for
you», all I did was to ask questions. Exactly *where* did I try to lie?

Best regards,

Jose Carlos Santos
.

Relevant Pages

  • JSH: Surrogate factoring and the k/T ratio
    ... addition needed mathematically by the concept of surrogate factoring. ... around how big k is relative to T--the k/T ratio. ... Posters can easily lie about math. ...
    (sci.math)
  • Surrogate factoring and the k/T ratio
    ... addition needed mathematically by the concept of surrogate factoring. ... around how big k is relative to T--the k/T ratio. ... Posters can easily lie about math. ...
    (sci.crypt)
  • Re: Surrogate factoring and the k/T ratio
    ... posters claiming it worked no better than random. ... around how big k is relative to T--the k/T ratio. ... at all--so I used three digit primes, ... but I did do other checks and saw it factoring rather ...
    (sci.crypt)
  • Re: Surrogate factoring, random is better?
    ... >> I keeps seeing posters claiming that my surrogate factoring method ... >> worse than even chance or at best only as good as a chance method ... >> roots constraint. ...
    (sci.crypt)
  • Re: Surrogate factoring, random is better?
    ... >> I keeps seeing posters claiming that my surrogate factoring method ... >> worse than even chance or at best only as good as a chance method ... >> roots constraint. ...
    (sci.math)