Software Demo 100 Bytes down to 7 Bytes Losslessly
Tim Bernard
approach to the problem.
I will more then happily demonstrate the software publiclly.
Heres how i will do it.
software application exists on cd rom 1
and is installed on notebook #1
anyone submits a cd rom that contains a 100 byte file.
Notebook 1 compresses data submitted from cd rom and
saves to floopy disk #1 a file that is 6 bytes
Floopy disk #1 is inserted to notebook #2 and uncompressed back to 100
bytes.
The entire setup can be inspected including the floopy disks at any time
during the demonstration.
I live just outside of Toronto Ontario and will demonstrate this software
this summer.
Eric Bodden
coming to Toronto for that...
Eric
Randall R Schulz
Should that work, it would be what we in the biz call a rigged demo.
Will it work with all 2^800 distinct 100-byte files? As has been
abundantly pointed out here, it will not. You can crunch numbers and
twiddle bits until god calls the cows home, but ten pounds will not fit
in a five-pound bag.
Among those reading here, only you seem to need to learn this subtle truth.
Perhaps you should go play in the highway.
Randall Schulz
Eric Bodden
about that?
I think we agree that I could give you 2^(800) pairwise different files of
length 100 byte, don't we?
That would be not less than 6.6680144328798542740798517907213e+240 different
ones.
Let's say you managed to "code" everyone of them, bijective, even if it
takes 1000 years, to 7 bytes each. Then you can NOT - you understand? - NOT
have more than 2^(8*7) = 2^56 = 72057594037927936 files!!!
So since approximately fantastillions of these "encoded files" are equal,
how do you wanna reconstruct all _different_ source files out of them?
Going to bed now.
Good night.
Eric
Tim Bernard
one urself and surrender it at the door once it works. :)
"Eric Bodden" <e...@ukc.ac.uk> wrote in message
news:b82270$eoi$1...@athena.ukc.ac.uk...
Tim Bernard
!
"Randall R Schulz" <rrsc...@cris.com> wrote in message
news:1%%oa.7572$JX2.4...@typhoon.sonic.net...
Tim Bernard
wrong.
"Eric Bodden" <e...@ukc.ac.uk> wrote in message
news:b822j3$eph$1...@athena.ukc.ac.uk...
Rolf Schroedter
with filenames of ~100 bytes ... ;-)
Rolf.
---------------------------------------------------------------
Rolf Schroedter, German Aerospace Center
Remove .nospam to reply: mailto:Rolf.Sc...@dlr.de.nospam
Salvador Fandiño García
> keep my notebooks if it doesn't work but at least have the balls to bring
> one urself and surrender it at the door once it works. :)
actually, I think that you should take one laptop and any other
guy we can trust another one, then you install the software in
both laptops and run the demo, if it works you get both laptops
and if not the other guy takes them, ok?
just because of wifi and things like that ;-)
- Salva
Ojala Pasi 'Albert'
Tim Bernard <notmy...@server.com> wrote:
>I will more then happily demonstrate the software publiclly.
>Notebook 1 compresses data submitted from cd rom and
>saves to floopy disk #1 a file that is 6 bytes
>
>Floopy disk #1 is inserted to notebook #2 and uncompressed back to 100
>bytes.
Why the floppy disk? 7 bytes is so little that you can type it in
from one machine to the another.
-Pasi
--
"Why do you love me?" "You know, you've always known. I wish I could be
you. I wish I could know the joy you know all the time." "And the pain,
you want that as well?" "Your pain? Certainly. I'll take your brand of pain
anytime, as they say." -- Lestat and Louis in "The Tale of the Body Thief"
Tim Bernard
"Ojala Pasi 'Albert'" <alb...@pikkukorppi.cs.tut.fi> wrote in message
news:b8373t$nf8$1...@news.cc.tut.fi...
Salvador Fandiño García
so you are definitively going wireless ? :-)
- Salva
Giovanni Motta
What's the compression on files of 50 bytes ?
And what about 25 bytes ?
G.
Kelsey Bjarnason
> u limit ur solutions to the problem by going about the problem completely
> wrong.
"Wrong"? What wrong? It's a very simple question: if n bits can only map
to 2^n files, how the hell do you plan to map _more_ than that using only
n bits?
Kelsey Bjarnason
> It will work with any 100 byte files. 10 poinds will fit into a 5 ounce bag
> !
Only if this is lossless compression, in which case, ho hum, who cares?
Eric Bodden
> to 2^n files, how the hell do you plan to map _more_ than that using only
> n bits?
I think a part of his brain got just lost in one of his n-th dimensions :-/
Teo van der Vlies
I can hardly wait to see a demo of your discovery.
I am afraid that you can not reach a lot of interrested people like me
because of the the travel costs and time to visit your demonstration.
Is it possible to send me the 7 bytes decoder, without losing the secret of
the brilliant encoder.
If I have the decoder, I can send you many 100 bytes files, that you have to
compress to 7 bytes files and send it back to me. ( You can also provide an
internet service doing the encoding of 100 bytes. )
When I run your decoder with a 7 byte file as input it will convert it back
to the 100 byte file I generated myself.
Teo.
"Tim Bernard" <notmy...@server.com> wrote in message
news:SG%oa.96731$BQi....@news04.bloor.is.net.cable.rogers.com...
Glen Smith
system
"Giovanni Motta" <g...@ieee.org> wrote in message
news:3EA55589...@ieee.org...
Glen Smith
Sorry.
Yes commercialization of the application i suspect would be via cpu
clustering of 64 cpus or such on faster system.
"Teo van der Vlies" <teo_...@hotmail.com> wrote in message
news:hKfpa.686012$sj7.27761064@Flipper...
Eric Bodden
1200
> system
Eric Bodden
> clustering of 64 cpus or such on faster system.
using a calculator first...
Teo van der Vlies
Sorry, I was talking to Tim Bernard.
Who are you by the way ?
Why you are answering the questions for Tim ?
Teo.
"Glen Smith" <gl...@smith.com> wrote in message
news:jMgpa.100864$BQi....@news04.bloor.is.net.cable.rogers.com...
Eric Bodden
Randall R Schulz
Teo van der Vlies wrote:
> Glen,
>
> Sorry, I was talking to Tim Bernard. Who are you by the way ? Why you
> are answering the questions for Tim ?
>
> Teo.
Teo,
Perhaps this offers an explanation for Tim / Glen's wonderous, magical
compression...
If Tim / Glen can fit two personalities in one brain, then probably the
world in which Tim / Glen lives admits posibilities not open to those of
us in the real world.
Or perhaps the fact that Tim and Glen post from the same host is just a
coincidence?
RRS
Tim Bernard
"Teo van der Vlies" <teo_...@hotmail.com> wrote in message
news:JFhpa.690366$sj7.27936133@Flipper...
Tim Bernard
"Eric Bodden" <e...@ukc.ac.uk> wrote in message
news:b84733$mug$1...@athena.ukc.ac.uk...
Tim Bernard
"Kevin Easton" <kevin@-nospam-pcug.org.au> wrote in message
news:newscache$hmyrdh$kmn$1...@tomato.pcug.org.au...
> Tim Bernard <no...@server.com> wrote:
> > Oh sorry my brother was using the comp and i didn't switch users sorry
>
> Your brother... who has a different last name? Well, I guess we already
> knew that logic wasn't your strong point.
>
> - Kevin.
>
lucas
"Kelsey Bjarnason" <kel...@xxnospamyy.telus.net> ha scritto nel messaggio
news:pan.2003.04.22....@xxnospamyy.telus.net...
A lot of time ago, there was a company claiming a similar compression; when
someone asked them how they can compress anything, against the basic laws of
information theory, they answered: "we simply do not use those laws".
In other words, I could claim that I can fly, because I am not using
Newton's gravity laws :) :P
Thomas Richter
>> "Wrong"? What wrong? It's a very simple question: if n bits can only map
>> to 2^n files, how the hell do you plan to map _more_ than that using only
>> n bits?
> A lot of time ago, there was a company claiming a similar compression; when
> someone asked them how they can compress anything, against the basic laws of
> information theory, they answered: "we simply do not use those laws".
> In other words, I could claim that I can fly, because I am not using
> Newton's gravity laws :) :P
Hmm. Intentionally forgot to pay your gravity bill that month? (-:
So long,
Thomas
Luigi
"Tim Bernard" <notmy...@server.com> ha scritto nel messaggio
Dale King
> Ok obviosuly ive attracted naysayers because most everyone has the wrong
> approach to the problem.
> I will more then happily demonstrate the software publiclly.
>
> Heres how i will do it.
I have another idea how you could demo it without exposing it to the world.
This method does not completely preclude the possibility of a rigged demo,
so it will not necessarily prove to any one that it does work. But if you
are actually sincere and think it does work it would allow everyone else to
prove to you that it does not work (which it of course cannot work).
It would be simpler in this case to use your claimed 64 bits to 35 bits
compression as that is a more manageable size chunk of data. You said this
runs in less than a second, so should be even faster on a server.
You put up a website form somewhere that lets you enter a 64-bit number
(i.e. only 16 hex characters). It uses CGI or servlets to run your code on
the server and spits out the 35-bit result. You also put up another page
that does the inverse where I can put in a 35-bit number and it gives me
back the 64-bit result. To make the test more believable those should not be
on the same server, preferably not even in the same country. It should be
pretty easy to automate code on our side that automates testing this.
As I said this does not eliminate the possibility of a rigged test. You
could just generate a random bit pattern and record the mapping between the
two. The two sites just looks up this mapping. But with enough testing you
will eventually run out of room for keeping all these mappings. And having
them on different servers makes this slightly more difficult.
--
Dale King
Vic Drastik
> it could be typed yes,
> > >saves to floopy disk #1 a file that is 6 bytes
> > >Floopy disk #1 is inserted to notebook #2 and uncompressed back to 100
> > >bytes.
> > Why the floppy disk? 7 bytes is so little that you can type it in
> > from one machine to the another.
> > -Pasi
This is interesting - what is the largest number of bytes that can be
losslessly compressed down to 3 bytes using your method. If 100 bytes can be
compressed to 7 bytes, I would guess the answer is 100**(3/7) = 7.197, or 7,
after rounding down. - am I right?
Vic
Glen Smith
"Vic Drastik" <12345...@bigpond.com> wrote in message
news:%CGpa.21226$1s1.3...@newsfeeds.bigpond.com...
Gib Bogle
> "Tim Bernard" <notmy...@server.com> wrote in message
> news:SG%oa.96731$BQi....@news04.bloor.is.net.cable.rogers.com...
>
>>Ok obviosuly ive attracted naysayers because most everyone has the wrong
>>approach to the problem.
>>I will more then happily demonstrate the software publiclly.
>>
>>Heres how i will do it.
>>
>
>
> I have another idea how you could demo it without exposing it to the world.
(snip)
If someone told you he could levitate, would you want to perform a test
to assess his credibility?
Gib
Dale King
news:3EA77621...@ihug.co.nz...
If he incessantly, annoyingly kept claiming to do the impossible, I would
eventually say prove it. Not to assess his credibility, but to discredit
him, to get him to shut up. Tim has no credibility since he is claiming
something that is provably impossible. But if we had a way to show him that
he is wrong perhaps we can get him to shut up.
I also think disproving him is more than just a personal thing, it is a
community service with legal implications. He may try to go get money from
people. If it can be shown that it was demonstrated that he was wrong and he
still took someone's money after that, then that strengthens a fraud case
against him. If he did not know that he was wrong, then it is a weaker case
for fraud against him.
--
Dale King
Gib Bogle
...
> I also think disproving him is more than just a personal thing, it is a
> community service with legal implications. He may try to go get money from
> people. If it can be shown that it was demonstrated that he was wrong and he
> still took someone's money after that, then that strengthens a fraud case
> against him. If he did not know that he was wrong, then it is a weaker case
> for fraud against him.
That is a good point.
Gib
Falk Hueffner
No, it's a bad point. Anybody investing money into a scam that is not
only obviously, but *mathematically provable* a scam, really deserves
to loose money.
(Other scamsters have done better than this... for example claiming to
be able to compress not all, but only "random" files is not provably
wrong because "random" is not well-defined.)
--
Falk
Gib Bogle
> No, it's a bad point. Anybody investing money into a scam that is not
> only obviously, but *mathematically provable* a scam, really deserves
> to loose money.
My mother used to say "It'd be a hard world if we all got our deserts".
Gib
Eric Bodden
Isn't desert that huge thing with the whole lot of sand in it?
Dale King
> > My mother used to say "It'd be a hard world if we all got our deserts".
>
> Isn't desert that huge thing with the whole lot of sand in it?
But it would be a hard world if we all got ours ;-)
--
Dale King
Gib Bogle
>>My mother used to say "It'd be a hard world if we all got our deserts".
>>
>
> Isn't desert that huge thing with the whole lot of sand in it?
Your "deserts" (emphasis on the second syllable) are what you deserve.
Maybe she said "just deserts" ("just" in the sense of justice). English
is a funny language.
Gib
Errol Smith
wrote:
>> My mother used to say "It'd be a hard world if we all got our deserts".
>
>Isn't desert that huge thing with the whole lot of sand in it?
>
I thought it was what you ate after dinner? Oh never mind..
Errol Smith
errol <at> ros (dot) com [period] au
Dale King
news:3eabc890...@news.ros.com.au...
> On Fri, 25 Apr 2003 08:35:55 +0100, "Eric Bodden" <e...@ukc.ac.uk>
> wrote:
>
> >> My mother used to say "It'd be a hard world if we all got our deserts".
> >
> >Isn't desert that huge thing with the whole lot of sand in it?
>
> I thought it was what you ate after dinner? Oh never mind..
No dessert is what you eat after dinner.
--
Dale King
Benny Baumann
And Tim might be the moderator of it ...
"Dale King" <Dale...@Thomson.net> wrote in news:3ea7...@news.tce.com...