Re: New form of recursive compression tested and proven. Now how do I market it?



<tom@xxxxxxx> wrote in message news:074e4133-fb24-42b2-82f0-6001fdb8ff33@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Jun 4, 8:02 am, "Paul" <p...@xxxxxxxxxxxx> wrote:
<t...@xxxxxxx> wrote in message

news:fd764921-31ce-4853-8c41-2a9c9a4396a1@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx



> On Jun 3, 5:10 pm, "Paul" <p...@xxxxxxxxxxxx> wrote:
>> <t...@xxxxxxx> wrote in message

>>news:a579407a-8d8f-43f3-a265-4715f7cb0b1d@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

>> > On Jun 3, 3:14 pm, "Paul" <p...@xxxxxxxxxxxx> wrote:
>> >> You got the wrong person. You must be referring to the person that
>> >> started
>> >> this thread. I never made such a claim.

>> >> Paul

>> > Ah, perhaps, you trolls are so interchangeable these days...

>> > Well since you're on a infinite compressor bender too, you can >> > answer
>> > the claim too.

>> > Tom

>> Tom, my claim is that I can recursively reduce random data. Does that
>> sound
>> impossible to you?

> You need stronger definitions of what you claim to do. Recursively
> could include the set of operations performed 0 times just as easily
> as N>0 times. You haven't defined what "random" data is either.

> If you're claiming that you can compress all N-bit strings (N>1) to a
> size of something less than N-bits long. You're a liar. It's trivial
> to disprove. You can't compress all 2-bit strings to 1 bit because 1
> bit can only represent half as many values as a 2-bit string. You
> can't compress all 3-bit strings to 2 or 1 bit because even combined
> (ignoring a need to be uniquely decodable) where are only 6 2- and 1-
> bit strings whereas there are 8 3-bit string. You can continue this
> argument on as N-bit and the sum of all possible 1,2,3,...,N-1 bit
> strings. \sum_{i=1}^{N-1}2^i will always be less than 2^N (hint: the
> sum is equal to 2^N - 2). Even if you somehow allowed in the zero-
> length string as a possible encoding, you're still only up to a sum of
> 2^N - 1. There are 2^N values you need to encode. So therefore you
> failed. And this of course is ignoring the fact you need some escape
> bits to tell one length of string from another.

> So now that I've disproven that ANY FORM of random data compressor is
> impossible, you're continuing to maintain that you've done it is just
> you being a troll. And there is no two ways about it (excuse the
> pun). You're now aware that it can't be done, so maintaining that it
> can is just you being contrary.

> What you need to do is stop trolling usenet because you're wasting
> your life. If you want to go out and have a good time hit the local
> pub, pound a few back and talk big with the folks at the counter. At
> least then you'll be interacting with real people in real life as
> opposed to getting a rise out of people remotely and anonymously.
> Honestly, I'm not saying this to be mean, you really ARE wasting
> everyones [including your own] time.

> Tom

Tom, I'm beginning to believe that there is no such thing as Random Data.
But I know the depth at which people here consider data Random so I go by
that. But for the sake of argument, let's call Random data whatever you
believe exhibits the greatest entropy.

There is no trolling going on here, Tom. You disbelieve so you consider
anyone that does believe to be trolling.

The point though is that it doesn't matter if the data is random or
not. messages of length M where M<N cannot represent all possible N-
bit messages. Whether they're random or enumerated. Like a program
that simply emits all 1024-bit integers will need to have at least
1024-bits of storage to represent the counter. Even though the
integers are sequential and totally non-random. So to have a program
that emits all N-bit messages, you'll need either N-bits [or more] of
input, or combine M<N bits of input with bits from the program.

And if you want to emit specific messages than the input must uniquely
determine the output which means you either can't compress all N-bit
messages, or you violate M>N on average.

Tom

Then by your understanding it should be impossible to save 12941 bytes of data from the million bit bin file, correct?

Paul

.

Relevant Pages

  • MORE on VH2
    ... Note VH2 is meant to be a bijective huffman file ... compressor from the set of all files to the set of all files. ... Keep in mind they work in the pure string space of all ...
    (comp.compression)
  • Re: Of interest to guitar players here ONLY (new handmade UK)
    ... Newtone strings are so good that you miss them when not there. ... it gives that 'string first' sound with the ever present string ... The blend of the three sound sources plus his rather expensive ... Also try the Bomb Factory compressor ...
    (rec.music.makers.songwriting)
  • Re: Fastest way to search a string for the occurance of a word??
    ... Performing thousands of tests over random data, ... You may have read somewhere that quicksort has Operformance in the ... > fairly early in the string, which is where String.IndexOf does better. ... > checking that this really is the performance bottleneck, ...
    (microsoft.public.dotnet.languages.csharp)
  • Re: New form of recursive compression tested and proven. Now how do I market it?
    ... bit can only represent half as many values as a 2-bit string. ... I'm beginning to believe that there is no such thing as Random Data. ... that simply emits all 1024-bit integers will need to have at least ... determine the output which means you either can't compress all N-bit ...
    (comp.compression)
  • Re: Compressing hash strings into two directions - hashing idea
    ... - Arihmetic compressor to the that string end with low ascii ... palindromes do no longer produce collisions. ... doesn't points 1-4 taken together mean that, when one string is ... Lets compress a 20 byte long hash-array 7 rounds in left-right ...
    (sci.crypt)