Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.compression > #720 > unrolled thread

I am taking Mark Nelsons challenge

Back to article view | Back to comp.compression

Contents

  I am taking Mark Nelsons challenge Michael <michaelhh@gmail.com> - 2012-01-06 02:10 -0800
    Re: I am taking Mark Nelsons challenge Michael <michaelhh@gmail.com> - 2012-01-06 09:29 -0800
      Re: I am taking Mark Nelsons challenge Earl_Colby_Pottinger <earlcolby.pottinger@sympatico.ca> - 2012-01-24 07:30 -0800
    Re: I am taking Mark Nelsons challenge Earl_Colby_Pottinger <earlcolby.pottinger@sympatico.ca> - 2012-01-06 09:46 -0800
      Re: I am taking Mark Nelsons challenge Ernst <Ernst_Berg@sbcglobal.net> - 2012-01-23 22:03 -0800
        Re: I am taking Mark Nelsons challenge Sebastian <s.gesemann@gmail.com> - 2012-01-24 01:45 -0800
          Re: I am taking Mark Nelsons challenge Ernst <Ernst_Berg@sbcglobal.net> - 2012-01-24 08:36 -0800
            Re: I am taking Mark Nelsons challenge pfraser <pete_fraser@comcast.net> - 2012-01-24 08:50 -0800
            Re: I am taking Mark Nelsons challenge Sebastian <s.gesemann@gmail.com> - 2012-01-25 04:07 -0800
        Re: I am taking Mark Nelsons challenge Earl_Colby_Pottinger <earlcolby.pottinger@sympatico.ca> - 2012-01-24 07:22 -0800
          Re: I am taking Mark Nelsons challenge Ernst <Ernst_Berg@sbcglobal.net> - 2012-01-24 09:07 -0800
            Re: I am taking Mark Nelsons challenge Earl_Colby_Pottinger <earlcolby.pottinger@sympatico.ca> - 2012-01-24 12:43 -0800
    Re: I am taking Mark Nelsons challenge jacko <jackokring@gmail.com> - 2012-01-24 12:48 -0800

#720 — I am taking Mark Nelsons challenge

I am doing it with math only... As I am not a software encoder.

I have a formula and description that I am willing to reveal to anyone
with a NDA

The initial version I will have in formula will show a 23% (rounded)
space savings.


Take a chance, sign a NDA... and run the math.


As part of this, I will require proof of personhood. Aka no false
identities. I wont lose this to a crook.

This can be a drivers license (You can cover the address portion if
desired), State ID, or Federal ID.

[toc] | [next] | [standalone]

#723

On Jan 6, 2:10 am, Michael <michae...@gmail.com> wrote:
> I am doing it with math only... As I am not a software encoder.
>
> I have a formula and description that I am willing to reveal to anyone
> with a NDA
>
> The initial version I will have in formula will show a 23% (rounded)
> space savings.
>
> Take a chance, sign a NDA... and run the math.
>
> As part of this, I will require proof of personhood. Aka no false
> identities. I wont lose this to a crook.
>
> This can be a drivers license (You can cover the address portion if
> desired), State ID, or Federal ID.

And thus a consequence for mixing two theories at once. Signal theory
and storage theory and oops... if I do this part it now requires this
value to change.

So where I thought I was spending four bit, I need 16. Since the gain
was four bits, this means I am negative eight. Even with a
modification to show I could get along with 12... this still leaves me
in the hole.

The only way this could be salvaged is if one quarter plus of the
results fall into proper play.

Sigh...

[toc] | [prev] | [next] | [standalone]

#815

On Jan 6, 12:29 pm, Michael <michae...@gmail.com> wrote:

> So where I thought I was spending four bit, I need 16. Since the gain
> was four bits, this means I am negative eight. Even with a
> modification to show I could get along with 12... this still leaves me
> in the hole.

That is my point in my other posts.

Instead of making your claim in the first place, would it not been
better do the testing in the first place and find your big error
first?

Second, if you are requiring a NDA then you are not going to find much
interest/help.  If instead you had posted your math theory in the
first place you would have been told your mistake very early on and
either not spent so much time on it or would be further along in
developing your theory.

PS. Please learn programming.

[toc] | [prev] | [next] | [standalone]

#724

On Jan 6, 5:10 am, Michael <michae...@gmail.com> wrote:
> I am doing it with math only... As I am not a software encoder.
>
> I have a formula and description that I am willing to reveal to anyone
> with a NDA
>
> The initial version I will have in formula will show a 23% (rounded)
> space savings.

NO SOFTWARE MEANS YOU GOT NOTHING.

Learn to write code, if that is too hard for you to do than it is
doubtful that you have to mental tools to develop a good compressor
using pure math.

And what the good is a NDA?  If the code can not be tested by the
public it is worthless junk.  In-case you are not aware of it EVERY
scam artist talks about NDAs. All the developers of lossless
compressors in use today have NOT asked for one!

A number of developers have kept their system secret by only shipping
compiled programs, The people behind RAR to my knowledge has never
released the code to the compressor but released the source code to
the decompressor. But the people who insist on NDAs have ZERO
marketshare. No-one wants to work for you for free, no-one wants a
bunch of empty claims.

Publish your methods or ship working code, anything less is garbage
because anyone can post empty claims.

[toc] | [prev] | [next] | [standalone]

#810

On Jan 6, 9:46 am, Earl_Colby_Pottinger
<earlcolby.pottin...@sympatico.ca> wrote:
> On Jan 6, 5:10 am, Michael <michae...@gmail.com> wrote:
>
> > I am doing it with math only... As I am not a software encoder.
>
> > I have a formula and description that I am willing to reveal to anyone
> > with a NDA
>
> > The initial version I will have in formula will show a 23% (rounded)
> > space savings.
>
> NO SOFTWARE MEANS YOU GOT NOTHING.
>
> Learn to write code, if that is too hard for you to do than it is
> doubtful that you have to mental tools to develop a good compressor
> using pure math.
>
> And what the good is a NDA?  If the code can not be tested by the
> public it is worthless junk.  In-case you are not aware of it EVERY
> scam artist talks about NDAs. All the developers of lossless
> compressors in use today have NOT asked for one!
>
> A number of developers have kept their system secret by only shipping
> compiled programs, The people behind RAR to my knowledge has never
> released the code to the compressor but released the source code to
> the decompressor. But the people who insist on NDAs have ZERO
> marketshare. No-one wants to work for you for free, no-one wants a
> bunch of empty claims.
>
> Publish your methods or ship working code, anything less is garbage
> because anyone can post empty claims.

Least we forget that lofty goals are reached with insightful dreams.

Who are we to kick ass on anyone that looks foreign to the established
guild.

@ Michael  Dare to dream!

@ everyone else.. No one shares things because it sucks to do so here.

It's honest.

[toc] | [prev] | [next] | [standalone]

#811

On 24 Jan., 07:03, Ernst wrote:
>
> @ Michael  Dare to dream!
> @ everyone else.. No one shares things because it sucks to do so here.
> It's honest.

That depends on what "things" are. If "things" are the typical data
compression crackpottery, the author will first be politly reminded of
the pointlessness of his goals and actions. If he/she continues to
believe in "things" that provably don't exist, refuses to accept
advice (from more knowledgable people) on how to improve his/her
understanding of the topic and posts more nonsense, he/she should be
prepared for a sh!tstorm. I think this applies to many other groups
with technical/scientific/mathematical topics. For example, go to a
physics group, claim to have developed a perpetual motion machine and
see what happens. For example, go to a math group, claim to have found
an integer larger than 2 but smaller than 3 and see what happens. For
example, go to a geodesic group, claim that the earth is flat disc and
see what happens...

[toc] | [prev] | [next] | [standalone]

#820

On Jan 24, 1:45 am, Sebastian <s.gesem...@gmail.com> wrote:
> On 24 Jan., 07:03, Ernst wrote:
>
>
>
> > @ Michael  Dare to dream!
> > @ everyone else.. No one shares things because it sucks to do so here.
> > It's honest.
>
> That depends on what "things" are. If "things" are the typical data
> compression crackpottery, the author will first be politly reminded of
> the pointlessness of his goals and actions. If he/she continues to
> believe in "things" that provably don't exist, refuses to accept
> advice (from more knowledgable people) on how to improve his/her
> understanding of the topic and posts more nonsense, he/she should be
> prepared for a sh!tstorm. I think this applies to many other groups
> with technical/scientific/mathematical topics. For example, go to a
> physics group, claim to have developed a perpetual motion machine and
> see what happens. For example, go to a math group, claim to have found
> an integer larger than 2 but smaller than 3 and see what happens. For
> example, go to a geodesic group, claim that the earth is flat disc and
> see what happens...

And what of the Man that invents anew?  Does he have to obey as well?

You guys have earned this from me.  On the one hand yes a lot of
blabbering goes on and that isn't always good on ether side and on the
other hand who knows how many people might have come up with something
new and this forum does it's best to bust that man's balls in the name
of conformity.

 For the record I have discovery.  I did it through the methods and
ways that are punished here.  I am working a modulus effort currently
on the million digit challenge.  A very respectable effort and yet I
discovered for myself how to factor numbers or not through modulus and
then discovered that folks have been fiddling with that for
centuries.  I know it has been written about but it's the human
experience that makes this worth doing and that you cannot beat out of
a man.

It's about a balance of input and output.  That the human mind is as
it is must be respected.  It's all we have.

For the record I am an ordinary home computer programmer and am self
taught on most things and I don't need the power trip when I am
interacting with the data compression folks.

Now let us move on to "you all know this or that can't be done"  None
of you know everything so never say never.

I am sure we are lacking experienced Quantum computer programmers here
so can any of you swear on their life that  Random Data will never be
compressed?

But for some reason this group almost hates anyone thinking
differently.

 I don't mind the dreamer as much as I mind the Mind-police here.  I
have been around this forum a decade now so I have a voice.

Be Well,

[toc] | [prev] | [next] | [standalone]

#821

Ernst wrote:
> On the one hand yes a lot of
> blabbering goes on and that isn't always good on ether side

"The only thing that really worried me was the ether. There is
nothing in the world more helpless and irresponsible and
depraved than a man in the depths of an ether binge, and I
knew we'd get into that rotten stuff pretty soon."

Raoul Duke -- Fear and Loathing in Las Vegas

[toc] | [prev] | [next] | [standalone]

#847

On 24 Jan., 17:36, Ernst wrote:
> On Jan 24, 1:45 am, Sebastian wrote:
> [...]
> Now let us move on to "you all know this or that can't be done"  None
> of you know everything so never say never.
> [...]
> But for some reason this group almost hates anyone thinking
> differently.

This is not true. Thinking differently is not necessarily a bad thing.
Talking bullsh!t is. There is a difference. If you want to be able to
make this kind of distinction successfully, you probably need to raise
your math skill level somehow.

> I don't mind the dreamer as much as I mind the Mind-police here.

Think whatever you like. Do whatever you like. But don't get pissed if
somebody calls you an idiot for believing in something that is as non-
existant as the solution for x of x+1=x.

> can any of you swear on their life that Random Data will never
> be compressed?

Depends on what you mean exactly. For example, a long string of zeros
can be the result of a random process. It might be very unlikely but
it is possible. In this case, the string can of course be represented
with less bits. But looking at examples is rather boring! Much more
interesting is  the following question: If we were to use a single
lossless compressor/decompressor pair that is built in advance and
repeat the random experiment many many times (that is, feeding the
compressor with different inputs from this random process), what is
the best possible _expected_ compression ratio? And even there, the
answer is: It depends. It depends on the actual random process. What
we do know with 100% certainty (no need to believe!) is that the best
possible _expected_ compression rate for a specific random process --
independent fair coin flips encoded as binary data -- is 1.0 which
means no compression on average is possible in this case! This is very
basic information theory stuff that is taught in introductory data
compression books.

http://en.wikipedia.org/wiki/Randomness
http://en.wikipedia.org/wiki/Random_process
http://en.wikipedia.org/wiki/Expected_value

What you should keep in mind: No sane person cares about compressing
one specific file. We want to be able to compress whole classes of
files with a single compressor/decompressor pair. Optimizing the
performance of such a compressor/decompressor pair boils down to
finding good statistical models for the kind of data you intent to
compress. If your statistical model is "independent fair coin flips"
there is PROVABLY (not to be confused with the word "probably")
nothing you can do about it in terms of lossless compression.

Cheers!
SG

[toc] | [prev] | [next] | [standalone]

#814

On Jan 24, 1:03 am, Ernst <Ernst_B...@sbcglobal.net> wrote:

> Least we forget that lofty goals are reached with insightful dreams.

And a solid foundation is needed to climb to those lofty goals. The OP
does not program code and does not plan to write code.

This is a problem, pure math is not working code.  And making pure
math work on a computer is not always easy.

Example, on most computers systems the answer to:
square_root(2.0) * square_root(2.0) does not equal to 2.0, it comes
very close but not exactly right.  This is a problem if you are
writting a lossless compression system.

> Who are we to kick ass on anyone that looks foreign to the established guild.

What guild? No-one is stopping them from doing whatever they want and
to give/sell it to anyone they want.
But when the OP says that they don't plan to do any coding, and the
original challenge requires a computer program to be met, guess what
work they are going to asking people here to do for them next?


> @ Michael  Dare to dream!

And learn to program.

> @ everyone else.. No one shares things because it sucks to do so here.

What sucks for them is having all their mistakes pointed out, and that
people here do not bow down worship their claims without proof.


As for real sharing, I have learnt lots of things from the people here
because I ask honest questions and ask for help as pointers to info
and URLs instead of expecting them to do all my work for me.

Too often people come here and elsewhere with lofty goals, but they
don't want to do the work needed to reach those goals, instead they
expect others to do the hard part for them.

The OP needs to learn programming or get a friend to do it for him,
not come here if he has no code.

PS. If he insists on pure math, where is the math? Worried people here
will point out the mistakes?

Maybe I seem too hostile, I hope not, I just want people to learn the
basics and do some testing before making their claims.

[toc] | [prev] | [next] | [standalone]

#822

On Jan 24, 7:22 am, Earl_Colby_Pottinger
<earlcolby.pottin...@sympatico.ca> wrote:
> On Jan 24, 1:03 am, Ernst <Ernst_B...@sbcglobal.net> wrote:
>
> > Least we forget that lofty goals are reached with insightful dreams.
>
> And a solid foundation is needed to climb to those lofty goals. The OP
> does not program code and does not plan to write code.
>
> This is a problem, pure math is not working code.  And making pure
> math work on a computer is not always easy.
>
> Example, on most computers systems the answer to:
> square_root(2.0) * square_root(2.0) does not equal to 2.0, it comes
> very close but not exactly right.  This is a problem if you are
> writting a lossless compression system.
>
> > Who are we to kick ass on anyone that looks foreign to the established guild.
>
> What guild? No-one is stopping them from doing whatever they want and
> to give/sell it to anyone they want.
> But when the OP says that they don't plan to do any coding, and the
> original challenge requires a computer program to be met, guess what
> work they are going to asking people here to do for them next?
>
> > @ Michael  Dare to dream!
>
> And learn to program.
>
> > @ everyone else.. No one shares things because it sucks to do so here.
>
> What sucks for them is having all their mistakes pointed out, and that
> people here do not bow down worship their claims without proof.
>
> As for real sharing, I have learnt lots of things from the people here
> because I ask honest questions and ask for help as pointers to info
> and URLs instead of expecting them to do all my work for me.
>
> Too often people come here and elsewhere with lofty goals, but they
> don't want to do the work needed to reach those goals, instead they
> expect others to do the hard part for them.
>
> The OP needs to learn programming or get a friend to do it for him,
> not come here if he has no code.
>
> PS. If he insists on pure math, where is the math? Worried people here
> will point out the mistakes?
>
> Maybe I seem too hostile, I hope not, I just want people to learn the
> basics and do some testing before making their claims.

Like I say it is a balance of input and output but it does suck here.
I hardly feel like it's a place for fun any more.
 I miss the tom-foolery we once had here. It was a fun place years
ago.  Now it's just fun for some.

 Just remember if I can discover new Maths anyone can including the
dreamers.

Have a fun song here guys > http://www.youtube.com/watch?v=jSFLZ-MzIhM

Here is another song ..  http://www.youtube.com/watch?v=qeJJOjb7fj4

Sometimes we are simply alone and lonely and I, for one, am willing to
be a friend to the alone and lonely.
That is something many of you fail at here,

[toc] | [prev] | [next] | [standalone]

#824

On Jan 24, 12:07 pm, Ernst <Ernst_B...@sbcglobal.net> wrote:

> Like I say it is a balance of input and output but it does suck here.

No, you did not. You want people to be believed on what they say
rather than on what they can prove.

> I hardly feel like it's a place for fun any more.
>  I miss the tom-foolery we once had here. It was a fun place years
> ago.  Now it's just fun for some.

Yes, it was fun years ago. And you know what you saw also during those
years?
1) Working Code.
2) Compression methods described in enough complete detail that you
could write working code.
3) New Data Structures designed help methods already designed to store
in more compact formats.

You know what you *DID NOT* see years ago when this place was such
fun?

1) People demanding NDAs be signed before you could see their
compression scheme.
2) People who would defend anyone who was not willing to show their
method/tests/coding first.
3) Claims of a secret guild just because we want to see proof of
claims first.

>  Just remember if I can discover new Maths anyone can including the dreamers.

But what has you new maths done that is useful?  If it can't do
something useful for compression, why bring it up.

And the rest of your post, like your new math has nothing to do with
getting compression working.

Want to make this place fun again? Lets talk about compression - real
working compression not vague pie-in-the-sky claims.

[toc] | [prev] | [next] | [standalone]

#825

You'll have a hard time. Forget about mapping big things to smaller things, and go for things of know size where you can change the bit pattern while maintaining constant size, such that change or no change has information content.

Cheers Jacko

P.S. The reals are countable. Don't believe the hype.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.compression

csiph-web