Ashley Labowitz <sporecoun
...@gmail.com> wrote:
> Don't forget that fractals have _unlimited_ self-similarity at smaller
> and smaller scales. Thus, the complexity for bits can always be
> borrowed, with promises to repay with information contained in not-yet-
> computed smaller scales (it's good enough for the the
> government* :) ). This algorithm is the key to "bit splicing" as I
> call it, evading the pigeonhole principle which only applies to finite
> sets (the fractal complexity is not even countable).
Very fine, but all your input data is finite by nature (the files on your harddisk)
and your output data is (hopefully) finite as well, so that buys you nothing.
And, of course the counting theorem applies for finite sets.
> I'd be surprised
> if even the most seasoned members of this group had heard of the
> Promise model of random compression.
What I don't get: Do you really know what you're talking about? You want
to make us believe that you've managed to get an understanding in fractals,
but yet deny to accept something as simple as the counting argument.
Or even elementary logic, if that counts. If you can compress every
file, then you can apply your compressor to its output immediatly again
and, by induction, arrive at a file that is - zero bits long. Thus,
effectively, you claim to be able to compress everything to nothing.
Well, I can do that as well:
rm *
Problem is: There is no decompressor.
> However, I found this technique too difficult to code in it's native
> form. In any case, due to the very nature of self-similarity (it
> being derived from a fixed equation after all), it's not clear how to
> use it to encode unlimited entropy. This required using the cyclic
> roots and rotation method.
Actually, are you just repeating some words you collected here and there,
or do you understand what you say? I suppose the former?
> Nevertheless, I now have the decompression tentatively resolved
> (fingers crossed).
No, you haven't.
> It's not coded up,
As I said, you haven't.
> but I can show that using
> Perfect numbers (Mersenne primes based on powers of two for the
> bifurcation) will recall the granularity level of the 'bit dust".
Are we playing a round of buzzword bingo today? A perfect number is a
number which is the sum of its proper divisors. For example, 6 is a
perfect number since 1+2+3 = 6. Primes are never perfect.
> Because of the Lyapunov chaos exponent of the polynomial in question,
> I have to multiply by 3.
Liapunov exponents are defines for dynamical systems, not for polynomials.
So what is your system? Why "3". huh?
> I'm calling these Gerfect numbers. I'm
> looking into some number theory textbooks from the library to find
> theorems relating to k-smooth numbers because the fractal dimensions
> are related to highly abundant numbers (not sure how, but I keep
> seeing it in the data).
Don't look into number theory textbooks. I recommend locking yourself into
your room, sit back for a day and think cleany on what you claim. It
doesn't require a math education to see the mistake.
Alternatively, three or for hints on the forehead should be applied.
As entertaining as it is, I guess it's time to stop this comedy. Otherwise,
I recommend to divide an angle by three with compass and ruler, find
the root of an arbitary 5th order polynomial or construct a square of
the area of a circle with ruler and compass, and post results here. It
will be also an enjoyable reading. (-:
So long,
Thomas
