Economic Impact of Iterative Random Compression
Plato777
Just as a mental exercise, suppose random compression were possible.
Just what impact would the existence of such an algorithm have on
our current world economic system. For starters, the existence of
libraries containing millions of square feet space of books along
with the librarians would be unnecessary. Historians would no
longer be needed to write books, since the common person could have
access to all raw historical data through indexes. The integrity of
all written and photographic data could be secured for millenia.
Imagine having a real picture of King Arthur, Julius Caesar, Socrates.
The actual progression of the sound and meaning of spoken language
could be mapped over time, making linguists and interpreters
unnecessary. Publishers and book-makers would be put out of
business. The perfect images of past best-selling books, modified
with modern words and phrases, would prevent authors from writing
new books, no economic incentive to do so. Future wars could be
prevented if an ancient verifiable compact between the parties
surfaced and settled disputes before any bloodshed happend.
Can anyone else think of other world-changing impacts if such an
algorithm existed?
Maybe this thread will convience anyone inventing an iterative-
random-compression algorithm to keep it concealed or maybe if the
benefits outway the liabilites to give it freely to humanity.
Willem
) Hi all;
)
) Just as a mental exercise, suppose random compression were possible.
) <snip>
)
) Can anyone else think of other world-changing impacts if such an
) algorithm existed?
Sure.
You will have proven that the second law of thermodynamics is in fact
false, which makes it possible to have unlimited energy.
I forgot what the URl is, but there's a description describing how random
compression leads to the falsehood of the second law.
By the way, someone posted a very very similar question not too long ago.
That person didn't get any serious answers either, so don't expect any.
SaSW, Willem
--
Disclaimer: I am in no way responsible for any of the statements
made in the above text. For all I know I might be
drugged or something..
No I'm not paranoid. You all think I'm paranoid, don't you !
#EOT
Marco Al
> Maybe this thread will convience anyone inventing an iterative-
> random-compression algorithm to keep it concealed or maybe if the
> benefits outway the liabilites to give it freely to humanity.
Or what you are really saying "Maybe you will get people to waste time".
Congratulations in my case you succeeded again, and maybe if Im crude enough
you will stop making up new nicks and go away?
You should not share your mental excercise with others, you cannot afford to
spare any.
Marco
Plato777
> Plato777 wrote:
> ) Hi all;
> )
> ) Just as a mental exercise, suppose random compression were possible.
> ) <snip>
> )
> ) Can anyone else think of other world-changing impacts if such an
> ) algorithm existed?
>
> Sure.
>
> You will have proven that the second law of thermodynamics is in fact
> false, which makes it possible to have unlimited energy.
What about Maxwell's daemon? There is always a special case.
I think the reversal of entropy is improbable, not impossible.
But of course the second law of thermodynamics is based on the
macroscopic view of our universe, not the quantum mechanic's view.
What did people say one-hundred years ago about the possiblity of
objects being suspended in air? (superconductivity at room temperature)
But the compression problem is not rooted to the physical
universe, it dwells in the abstract-logic plane of existence.
Take the number Pi for example. The digit formula for Pi in binary was not
found until the mid-1990's by some French guy. The formula was
overlooked by mathematical heavyweights in the past and most gave
up on the possibility of a formula. Pascal's Theorem was just proven
recently to the astonishment of many. The iterative compression is
mathematical not physical in nature; therefore, it could possibly
exist.
>
> I forgot what the URl is, but there's a description describing how random
> compression leads to the falsehood of the second law.
>
> By the way, someone posted a very very similar question not too long ago.
> That person didn't get any serious answers either, so don't expect any.
Yeah I know. I was trying to establish a reductio ad absurdum
arugment against the possibility of the algorithm's existence.
The hostility people show for its very existence causes it not to exist,
rightly so--we are not ready as a species to receive it.
Some people say it is not worth anyone's time to discuss the possiblity,
that it is pointless. I say that 90% of life is pointless. But
this discussion is more worthwhile than some other pointless persuits.
Erik.
Willem
)> ) Hi all;
)> )
)> ) Just as a mental exercise, suppose random compression were possible.
)> )
)> ) Can anyone else think of other world-changing impacts if such an
)>
)> Sure.
)>
)> You will have proven that the second law of thermodynamics is in fact
)> false, which makes it possible to have unlimited energy.
)
) What about Maxwell's daemon? There is always a special case.
) <snip>
) recently to the astonishment of many. The iterative compression is
) mathematical not physical in nature; therefore, it could possibly
) exist.
We've had all these arguments before.
If you want to troll successfully, you should at least try to come up with
new handwaving arguments.
) that it is pointless. I say that 90% of life is pointless. But
) this discussion is more worthwhile than some other pointless persuits.
It's not pointless to educate people on the mathematical impossibility of
something, so they can turn their energies to more useful pursuits.
Alex Sisson
everyone would live in harmony with peace and love and unity and respect etc.
Guenther von Knakspott
> recently to the astonishment of many.
which one?
> The iterative compression is
> mathematical not physical in nature; therefore, it could possibly
> exist.
You got things mixed up here. Precisely because it is mathematical we
can know that it is impossible. For physical phenomena, we asume that
the laws known to us are is universal in the sense that they apply
always and everywhere. However we can only asume so and never be 100%
sure. The laws might be proven to be wrong, or exceptions may be found
to them. This is so because the physical laws we know, are
generalisations, and parts of a model. Whatever certainty of their
validity we have, is derived from observations, experiments,
reproducibility, aka the scientific method. On the other hand,
mathematics is a formal system defined within strict rules. If a
mathematic theorem can be proved to be true then it cannot be proven
to be wrong, ever. There are enough mathematic proofs that "iterative
compression" as you call it is impossible. Mathematically there is no
way around these proofs. If you really are interested in this field,
you would do well in learning mathematics until you understand this.
Only then will you really be able to make progress.
> Yeah I know. I was trying to establish a reductio ad absurdum
> arugment against the possibility of the algorithm's existence.
> The hostility people show for its very existence causes it not to exist,
> rightly so--we are not ready as a species to receive it.
No, it is not the hostility that causes it not to exist. It just
doesn't period. The hostility comes from the fact that people who do
not take the trouble of learning enough mathematics or information
theory continually challenge proven facts out of greed for money or
fame, self delusion or plain stupidity. Their attitude often follows
this thought pattern: a) In the past there were brave men and women
who challenged established knowledge and at great risk proved they
were right. b) I have an idea which challenges established knowledge,
therefore I am brave and therefore I can prove I am right.
Unfortunately things don't work that way. Neither challenging the
establishment nor being brave are enough. First of all you have to be
right in the first place, and second you often have to put a lot of
work into proving it (ie. gathering enough knowledge through study to
be at least at the same height of that which you are challenging)
> Some people say it is not worth anyone's time to discuss the possiblity,
> that it is pointless. I say that 90% of life is pointless. But
> this discussion is more worthwhile than some other pointless persuits.
>
>
> Erik.
That is really sad, but if your life is really 90% pointless then you
might as well waste it as you will, at least you are entitled to that.
Just don't expect others to waste their lives along with yours.
regards
Plato777
> >
> > Can anyone else think of other world-changing impacts if such an
> > algorithm existed?
> >
> > Maybe this thread will convience anyone inventing an iterative-
> > random-compression algorithm to keep it concealed or maybe if the
> > benefits outway the liabilites to give it freely to humanity.
>
> everyone would live in harmony with peace and love and unity and respect etc.
Do you really believe that an utopia could result? Do you think
the free and open distrubution of information could have a "Pandora's Box"
effect?
I'm debating on whether I should share my compression findings with others.
The rudeness of others on this newsgroup, attacking the mere mention
of the possibility of iterative compression confuses me. Maybe they
were investor's burned in the past and stauntly refuse to entertain
the idea again.
If I do share my findings with this newsgroup I share ten years of work
with humanity with nothing accepted in return but credit for my work.
My question is that if I give my work freely to Humanity will enough
people stand with me if others try to steal my ideas and profit from
them. I believe the time for privileged information should end.
If I do decide to share my algorithm, I will force others through its
construction step by step and share the tedium and wonderment that I
have been through.
The main overall process of the algorithm follows what I call
the "Down Pillow" effect. Think about it.
I know what you all are thinking: "Look, another nut to debunk and
insult and to declare 'unclean'."
I ask only for patience. All pharisaical decriers please post
your hatred to another thread. You can flame me all you want
there. But remember your rants are a part of the public record.
Erik.
Look at that idiot throwing pieces of paper into the air, he should
be enjoying the show;-)
Willem
) Alex Sisson wrote:
)> everyone would live in harmony with peace and love and unity and respect etc.
)
) Do you really believe that an utopia could result? Do you think
) the free and open distrubution of information could have a "Pandora's Box"
) effect?
No, he was being sarcastic.
And although I think you're trolling, here is something to think about:
What do all the following sentences have in common:
It is impossible to find two integers a and b so that a/b = Pi.
It is impossible to find a 1:1 mapping between the real numbers and N.
It is impossible to find a prime larger than 5 that is divisible by 5.
It is impossible to compress random data.
Matt Mahoney
> ) Hi all;
> )
> ) Just as a mental exercise, suppose random compression were possible.
> ) <snip>
> )
> ) Can anyone else think of other world-changing impacts if such an
> ) algorithm existed?
>
> Sure.
>
> You will have proven that the second law of thermodynamics is in fact
> false, which makes it possible to have unlimited energy.
>
> I forgot what the URl is, but there's a description describing how random
> compression leads to the falsehood of the second law.
The proof is simple. From logic, we have
not P => (P => Q)
where P = "random compression is possible" and Q is whatever you want.
-- Matt Mahoney
Plato777
>
> No, he was being sarcastic.
>
> And although I think you're trolling, here is something to think about:
>
> What do all the following sentences have in common:
>
> It is impossible to find two integers a and b so that a/b = Pi.
ERROR.
Error of category: rational numbers and trancendental numbers don't
mix.
> It is impossible to find a 1:1 mapping between the real numbers and N.
ERROR.
Error of category: N and Z and Q are all SUBsets of R -- by
definition,
the quantity of elements of N have to be less the R. 1:1 between sets
must have as a pre-requirement that equal number of elements appear in
each set.
> It is impossible to find a prime larger than 5 that is divisible by 5.
ERROR.
Error of definition: an integer is divisible only by one and itself.
> It is impossible to compress random data.
ERROR -- if all statements have one thing in common.
The previous statements have one point in common: misuse of the
arbitary definitions of the words and elements used to compose the
statements.
Perhaps clarification of your definitons of the words used in the last
statement might allow common understanding of your meaning.
>
>
> SaSW, Willem
Plato777
I have mistakenly used "integer" when I meant "Prime" in my previous post.
Please excuse my ERROR.
Erik.
Severian
wrote:
>Willem <wil...@stack.nl> wrote in message news:<slrnc0gqpp...@toad.stack.nl>...
>
>>
>> No, he was being sarcastic.
>>
>> And although I think you're trolling, here is something to think about:
>>
>> What do all the following sentences have in common:
>>
>> It is impossible to find two integers a and b so that a/b = Pi.
>
>ERROR.
>Error of category: rational numbers and trancendental numbers don't
>mix.
Chef sings, "Pig and Elephant DNA Just Don't Mix."
What? Rationals and transcendentals don't intersect (his statement was
an instance of this), but they certainly "mix." pi/pi = 1.
>> It is impossible to find a 1:1 mapping between the real numbers and N.
>
>ERROR.
>Error of category: N and Z and Q are all SUBsets of R -- by
>definition,
>the quantity of elements of N have to be less the R. 1:1 between sets
>must have as a pre-requirement that equal number of elements appear in
>each set.
Horseshit. Are you from the 19th century? Set theory has come a long
way.
The set of even numbers is a subset of the set of integers. However,
they have an equal number of elements (cardinality).
The proof that the cardinality of R is greater than that of N requires
a more elegant argument. Look it up.
>> It is impossible to find a prime larger than 5 that is divisible by 5.
>
>ERROR.
>Error of definition: a -n-integer- prime is divisible only by one and itself.
The definition is what makes the statement true. How is that an error?
>> It is impossible to compress random data.
>
>ERROR -- if all statements have one thing in common.
>The previous statements have one point in common: misuse of the
>arbitary definitions of the words and elements used to compose the
>statements.
>
>Perhaps clarification of your definitons of the words used in the last
>statement might allow common understanding of your meaning.
I would advise you to learn some math. Or at least google for
"pigeonhole principle" and "counting argument." You might also want to
study up on Cantor, set theory, number theory, analogy, Zeosync and
Jules Gilbert.
- Sev
Phil Carmody
> Willem <wil...@stack.nl> wrote in message news:<slrnc0gqpp...@toad.stack.nl>...
>
> >
> > No, he was being sarcastic.
> >
> > And although I think you're trolling, here is something to think about:
> >
> > What do all the following sentences have in common:
> >
> > It is impossible to find two integers a and b so that a/b = Pi.
>
> ERROR.
> Error of category: rational numbers and trancendental numbers don't
> mix.
Do what? What's Pi/2? See, they do mix.
> > It is impossible to find a 1:1 mapping between the real numbers and N.
>
> ERROR.
> Error of category: N and Z and Q are all SUBsets of R -- by
> definition,
> the quantity of elements of N have to be less the R. 1:1 between sets
> must have as a pre-requirement that equal number of elements appear in
> each set.
Do what? N and Z and Q aren't necessarily subsets of R. In most
standard set-theoretic constructions of R (e.g. Dedekind's or Cauchy's)
they aren't.
> > It is impossible to find a prime larger than 5 that is divisible by 5.
>
> ERROR.
> Error of definition: an integer is divisible only by one and itself.
And -1 and -1*itself.
(yes I'm being pedanantic for that one)
> > It is impossible to compress random data.
>
> ERROR -- if all statements have one thing in common.
> The previous statements have one point in common: misuse of the
> arbitary definitions of the words and elements used to compose the
> statements.
>
> Perhaps clarification of your definitons of the words used in the last
> statement might allow common understanding of your meaning.
The last one is somewhat different from the previous 3.
It is possible to compress some data from a random source
(it's _not_ the data that's random, it's the source that's
random), just a negligibly small proportion of the time.
If your random number generator is incapable of giving you
0 100 times in a row, then get a better generator. If it
does it more than ~2^-100 of the time, then likewise get
a better generator.
Phil
--
Unpatched IE vulnerability: window.open search injection
Description: cross-domain scripting, cookie/data/identity theft, command execution
Reference: http://safecenter.net/liudieyu/WsFakeSrc/WsFakeSrc-Content.HTM
Exploit: http://safecenter.net/liudieyu/WsFakeSrc/WsFakeSrc-MyPage.htm
Sebastian
> ERROR.
> Error of category: N and Z and Q are all SUBsets of R -- by
> definition,
> the quantity of elements of N have to be less the R. 1:1 between sets
> must have as a pre-requirement that equal number of elements appear in
> each set.
ROTFL!!!
Besides that by typical definitions N Z & Q are not substets of R, only that
some elements of are can bte treated equal to elements of N, Z or Q for
sake of convenience, then |N| = |Z| = |Q| = number of even numbers = number
of all primes, etc. put all are less than |R| which "by accident" is equal
to |2^N| i.e. number of all subsets of N.
Learn more before starting trolling.
Sebastian Kaliszewski
--
"Never underestimate the power of human stupidity" -- from Notebooks of L.L.
Willem
)> It is impossible to find two integers a and b so that a/b = Pi.
)
) ERROR.
) Error of category: rational numbers and trancendental numbers don't
) mix.
It's 'transcendental', and Pi/Pi = 1, so they do mix.
They did actually have to *prove* that Pi isn't rational, you know, they
didn't just define it that way.
Are you familiar with the proof that sqrt(2) is not rational ?
)> It is impossible to find a 1:1 mapping between the real numbers and N.
)
) ERROR.
) Error of category: N and Z and Q are all SUBsets of R -- by
) definition,
) the quantity of elements of N have to be less the R. 1:1 between sets
) must have as a pre-requirement that equal number of elements appear in
) each set.
That's total nonsense.
There *is* a 1:1 mapping between the rational numbers and N.
And they actually had to *prove* that there is no 1:1 mapping between
N and R, you know, they didn't just define it that way.
Are you familiar with the proof there is no 1:1 mapping between N and R ?
)> It is impossible to find a prime larger than 5 that is divisible by 5.
)
) ERROR.
) Error of definition: an integer is divisible only by one and itself.
That was the point, yes. And this one is the odd one out. The other three
statements have been proven at one point or another, this one is true by
simple definition.
)> It is impossible to compress random data.
)
) ERROR -- if all statements have one thing in common.
) The previous statements have one point in common: misuse of the
) arbitary definitions of the words and elements used to compose the
) statements.
Nope, all my statements were perfectly valid. It's your understanding of
them that's lacking. You don't know much about mathematics, do you ?
Or, of course, you're trolling.
Are you familiar with the proof that you can't compress all data ?
) Perhaps clarification of your definitons of the words used in the last
) statement might allow common understanding of your meaning.
Okay.
I used the standard mathematical definition of all mathematical terms.
By 'to compress random data' I mean
'given a certain amount of any data, to be able to compress it'.
George Johnson
news:aa7c3e68.0401...@posting.google.com...
Think of it this way. This newsgroup exists for the free collaboration
of ideas.
The problem is that sharing ideas also can result in corruption of those
ideas and theft of those ideas.
Take for example all of the bullshit that SCO has unleashed on the LINUX
front.
Be cautious on anything you share. It is logical business sense.
Any generosity can be regarded as an opportunity to be screwed over.
However, the "perish or publish" mantra still has truth to it.
Any idea worth sharing, thusly not clearly shared, can be lost if the
author of that idea dies as all ideas die unless shared.
Dale King
>
> ERROR -- if all statements have one thing in common.
> The previous statements have one point in common: misuse of the
> arbitary definitions of the words and elements used to compose the
> statements.
>
> Perhaps clarification of your definitons of the words used in the last
> statement might allow common understanding of your meaning.
I agree that this statement is troublesome, not because Willem is wrong, but
by the fact that there are multiple definitions of the terms used in the
statement. There is the notion of randomness from Kolmogorov, where that
statement is true by definition. There is also the notion of a random
source. It is actually possible that you can compress data from a random
source because it is possible for the source to put out 1000 zeroes. It just
isn't very likely and the odds make it very low probability of payoff.
We also need to insert the word losslessly in the statement as well. I know
that is what Willem meant, but you have to be explicit.
I prefer the statement that it is impossible to *losslessly* compress all
possible inputs of size N bits by at least one bit. There is no
misinterpretation of that and it is easily proven by the counting argument.
--
Dale King
Ryan
><blah blah blah a whole bunch of stuff>
> I prefer the statement that it is impossible to *losslessly* compress all
> possible inputs of size N bits by at least one bit. There is no
> misinterpretation of that and it is easily proven by the counting
argument.
> --
> Dale King
I've got my Nomex(r) suit on, so go ahead and flame away, but here's my
$0.02...
$0.01: It is not "impossible to losslessly compress all possible inputs of
size N bits by at least one bit", it is only impossible to do it with a
single algorithm. Even with multiple algorithms it is impossible, but only
if you count the algorithm used as at least one bit. If you get to name your
files with .Z/.zip/.rar/etc. you are encoding information about the
compressor used, and that officially should be counted as a bit (or more,
ceil(log2(N)) bits for N compressors), but in real life it's assumed that a
single compressor and decompressor are a known pair, so you might gain a bit
or two overall.
$0.02: Does anyone actually >care< about compressing every possible file of
size > 16 bytes, random or otherwise? There aren't likely more than 2^256
files (or 2^1024 and 128 bytes, take your pick) that anyone has created,
even in the entire history of the universe. What do you care if you can't
compress 99.9999999999% of them as long as the important ones compress to a
size of your liking? You only care that your own pigeons have holes, not
that every possible pigeon-mutation that will ever be born until the end of
time can fit into your roost (or whatever it is that actually contains the
pigeon holes... does someone know, just FMI?).
Also, nobody seems to answer the basic question... how would the world
change if you could store EVERYTHING on a single CD?
My answer would be "not a whole lot", just because the access to such
information would be the hard part, figuring out which one of those 2^1024
files you actually need is what really matters. Yes, you could download the
entire contents of the internet onto a single computer, but how often do you
search for something, even amongst Google's 3,307,998,701 pages, and not
find what you're looking for? The original post said something about
indexes (or indices) being the way everything is accessed, but those indexes
would then become the actual information that everyone wanted, and so the
ultimate compression would be moot, these indexes would be the things that
were traded instead of individual files.
Also the elimination of the "interpreters", the writers, historians,
linguists and other documenters couldn't happen. People still need to
create content, go scan in pictures of Julius Caesar/etc. and interpret
what's going on. "The interest rate on XYZ bonds increased 0.3% last month"
is great to know, but some financial analyst has to tell the rest of the
world what that actually means. Also, advanced as computer systems are
these days, adding the sum total of human knowledge to someone's desktop
would just make for more headaches. How much trouble is it to get your
mothers to use e-mail, let alone unleashing the universe on them? More
people would >become< librarians, they would be the interpreters or
searchers, the go-betweens for the average human to turn such boundless data
into useful information. They would be the new oracles, the new prophets.
My firm belief is that books (and therefore authors and publishers) will
never disappear. I am a true geek, I love Project Gutenberg enough to cron
rsync it. I want to automate everything, even the stuff that doesn't need
to be automated, but I also have a house full of books. They are tangible,
and they are comfortable. Sometimes they are even faster than a search
engine if I don't know exactly what I'm looking for.
There's more to be said on this, but it's late and i'm fading fast, so that
is that.
I'm not actually anything (mathematician/philosopher/psychic/etc.), so
nothing here actually counts for anything, so think what you like.
- Ryan
Guenther von Knakspott
> size N bits by at least one bit", it is only impossible to do it with a
> single algorithm.
Nonsense. Choosing the proper algorithm for compression/decompression
is an algorithm in itself, and grouping this with the actual
compressors/decompressors you obtain a single algorithm again. Youre
Back to sqare 1.
[bla bla bla]
>
> Also, nobody seems to answer the basic question... how would the world
> change if you could store EVERYTHING on a single CD?
What if the moon where made of chocolate cake? See anyone giving an
answer to that stupid question? No? How come? I'll tell you, because
It is a waste of time to answer to every crackpot idea that gets
posted.
> My answer would be "not a whole lot",
[bla bla bla]
A few trivially obvious items.
.- you could reduce your networking cost by an arbitrary amount.
.- no one would ever need to buy a harddisk again, when it got full
you would simply compress the contents and free up new space.
.- you could transmit any type of content at any desired rate through
the internet.
etc. etc. etc.
That is why these idiots keep comming back all the time, because they
think they can get rich by sheer stupidity and lazyness.
> I'm not actually anything (mathematician/philosopher/psychic/etc.), so
So how come you have to go and post your blatantly unqualified
opinion?
Ryan
> > $0.01: It is not "impossible to losslessly compress all possible inputs
of
> > size N bits by at least one bit", it is only impossible to do it with a
> > single algorithm.
>
> Nonsense. Choosing the proper algorithm for compression/decompression
> is an algorithm in itself, and grouping this with the actual
> compressors/decompressors you obtain a single algorithm again. Youre
> Back to sqare 1.
I don't think i'm explaining this properly... I only mean that any one
program (zip, compress, rar, etc.) will encounter some files which compress
and some files which expand. Is there a mathematical reason to believe
that, given enough such programs and iterations, any file could not be
compressed? You _aren't_ actually defeating the counting argument because
you add information about the decompressor to a different place, it's not in
the file it's in the filename. This is similar to the BARF compressor
(http://www.cs.fit.edu/~mmahoney/compression/barf.html), it does compress
the file but it adds the information onto the filename, if you change the
archive name to bob.zip your entire file will be gone. The same principle
holds true in the real world, if you change your filename from bob.bz2 to
bob.zip, the ZIP decompressor has no idea what to do with a bzip2 compressed
file. Also in the real world this is a stupid idea, it takes far too much
time to try options on a single compressor let alone N different compressors
recursively, but is there any information-theory reason it couldn't work?
>
> [bla bla bla]
> >
> > Also, nobody seems to answer the basic question... how would the world
> > change if you could store EVERYTHING on a single CD?
>
> What if the moon where made of chocolate cake? See anyone giving an
> answer to that stupid question? No? How come? I'll tell you, because
> It is a waste of time to answer to every crackpot idea that gets
> posted.
>
ouch. Point taken
> > My answer would be "not a whole lot",
> [bla bla bla]
>
> A few trivially obvious items.
>
> .- you could reduce your networking cost by an arbitrary amount.
> .- no one would ever need to buy a harddisk again, when it got full
> you would simply compress the contents and free up new space.
> .- you could transmit any type of content at any desired rate through
> the internet.
> etc. etc. etc.
>
> That is why these idiots keep comming back all the time, because they
> think they can get rich by sheer stupidity and lazyness.
I thought it was brute force and ignorance, but either way it's the same...
>
> > I'm not actually anything (mathematician/philosopher/psychic/etc.), so
>
> So how come you have to go and post your blatantly unqualified
> opinion?
Because I thought through what I was saying, and it sounded logical, even
after I read it back. Also, from the formal training (6 years of schooling
in CS/related matter) i've had on the aforementioned subjects (excluding
psychogenics) as well as past experience (20 years of computer use, 18 of
programming), I felt qualified enough to share my opinion. Another reason
to do it is just to piss off the world, and how could I pass on such an
exciting proposition?
Other than "arroganter kerl" I couldn't find exactly who you are, how are
you likewise qualified to express your opinions?
Just FYI, http://www.m-w.com/cgi-bin/dictionary?va=guttural
Willem
) I don't think i'm explaining this properly... I only mean that any one
) program (zip, compress, rar, etc.) will encounter some files which compress
) and some files which expand. Is there a mathematical reason to believe
) that, given enough such programs and iterations, any file could not be
) compressed?
Yes. Almost all programs include enough header information so that their
contents can be recognized without use of the file name. The only ones
that do not are (afaik) those that are used with an encryption algorithm.
A second reason is that all 'real' compression programs target files that
are in some way redundant, and will therefore only compress files that have
redundancy in them.
Guenther von Knakspott
> recursively, but is there any information-theory reason it couldn't work?
Look, I'll try it the civil way one single time (I got this strage
feeling this morning, I think its called good humour):
You cannot compress every string no matter what you do, because there
are more strings of lenght n than of length m for m<n. For binary
strings there are 2^n strings of length n and a total of (2^n)-2
strings of a shorter length. For strings over larger alphabets the
difference actually grows with n.
So you cannot have a one to one relationship between the set of
strings of length n and the set of all strings of length smaller than
n (shorter strings).
Do you understand that?
Well, there is no way around that, no matter what you do.
Do you understand that too?
If you can't say yes to those two questions, go back and study some
literature on the matter, until you can.
I you come back and say "but couldn't whe just ...." you won't get any
civil answer from me anymore.
[snip]
> > > I'm not actually anything (mathematician/philosopher/psychic/etc.), so
> >
> > So how come you have to go and post your blatantly unqualified
> > opinion?
>
> Because I thought through what I was saying, and it sounded logical, even
> after I read it back. Also, from the formal training (6 years of schooling
Most of the people who are wrong, think through what they are saying,
and it "sounds logical" to them. Still they are and remain wrong. You
seem to be someone who should know better, otherwise I wouldn't be
giving you this answer. So you knew you where unqualified, you knew
you were saying somthing contrary to facts that have been explained
many times in this group, but instead of making an effort to
understand better, you go and share with the world. We are mired in
false information because virtually anyone can inflict his stupid
opinion on the world. Personally I am fed up with having to wallow
through crap all the time.
> in CS/related matter) i've had on the aforementioned subjects (excluding
> psychogenics) as well as past experience (20 years of computer use, 18 of
> programming), I felt qualified enough to share my opinion.
Sorry, you didn't feel so. Your posting was more like "I don't have a
clue about what I'm saying, but I can't help making a fool of myself
in front of you all" These were your actual words.
"I'm not actually anything (mathematician/philosopher/psychic/etc.),
so
nothing here actually counts for anything, so think what you like."
There also was something about being on the verge of falling asleep.
So in my opinion you were just asking for it.
> Other than "arroganter kerl" I couldn't find exactly who you are, how are
> you likewise qualified to express your opinions?
I wasn't flaming you for being unqualified, but for posting crap. As I
said before, It was yourself who explicitly pointed out your own
"un-qualification". That just helped fuel the flames.
> Just FYI, http://www.m-w.com/cgi-bin/dictionary?va=guttural
"GUTTERAL" the word he was using was "GUTTERAL", you wont find that in
Merriam-Webster. You have to read things properly. It will not have
bee lost upon you, that he was making very denigrating remarks about
the usage of language among certain ethnic groups, so I pointed out to
him that he was misusing the english language himself by
misspelling.You have to read things properly
regards
Jonathan Hoyle
<stupidity snipped>
> Can anyone else think of other world-changing impacts if such an
> algorithm existed?
Let's see...I can think of the following:
1. Pigs would fly
2. Snowstorm in hell
3. Stupid, pointless posts like this would disappear from the internet
Jonathan Hoyle
Ryan
completely disagreeing?
"Guenther von Knakspott" wrote...
> You cannot compress every string no matter what you do, because there
> are more strings of lenght n than of length m for m<n. For binary
<etc.>
> So you cannot have a one to one relationship between the set of
> strings of length n and the set of all strings of length smaller than
> n (shorter strings).
<etc.>
Quotes used liberally to denote "not really" status, compressed below means
pseudo-compressed
File of length N
_Filename_ of length M
Original file _total data_ I = N + M
"compressed" file of length N-1
"compressed" filename of length M+X, X>=1
"compressed" file total data J = (N-1) + (M+X)
File length of "compressed" file is (N-1) < N, therefore it is compressed.
In the "compressed" file, there are 1/256th the number of possible messages
as in the original, without further information there are 256 possible files
this could decompress to. Such further information is contained in the
filename, see the aforementioned BARF compressor for a working example.
Although the file itself is smaller, the overall information density (not
sure if this is the right term, ) of the "compressed" file is less than the
original file. To transmit it somewhere it will take MORE bits (J > I). It
doesn't matter that your 1MB file shows up compressed as 1 byte, you now
have a 1MB+ filename associated with it which can't be changed without
losing information. It doesn't take anything away, it just moves where it's
keeping the information. Since filenames generally have character-set
limitations, it's reasonable to say that J will be significantly larger than
I, since the code size is reduced to the specified character set as opposed
to all 256 ASCII characters. For example, for ISO-9660 filenames the
filename length would increase approximately 8 bytes for ever 5 bytes of
file data (-2.83 bits per byte). In this case, I am ignoring the fact that
they are limited to 8.3 characters in total, but that would just save you
the trouble of doing such a stupid compression since you can't "compress" by
more than 5 bytes MAXIMUM using this method.
> I you come back and say "but couldn't whe just ...." you won't get any
> civil answer from me anymore.
Personally I don't think your original reply was particularly civil, but
i'll leave it at this: yes i'm a dick, it's a curse I have to live with.
Unfortunately, so do you, at least until you click Next.
Guenther von Knakspott
> this could decompress to. Such further information is contained in the
> filename, see the aforementioned BARF compressor for a working example.
What are you talking about? The filename is part of the file. You
cannot achieve compression that way. Look, BARF is a joke. I guess
Matt Mahoney called it that way after the Baruff-Hogsbrenn device. But
I guess you haven't heard about that either. Back in 79 these two guys
Jonathan Baruff and Richard Hogsrbrenn showed up with a device that
was able to compress virtually any kind of information stream. The
device was tested by a board including members of the NSA. It passed
every test and caused a lot of exitement. Then the Reagan
administration came to power. The NSA got agressive and put a lot of
pressure on Baruff and Hogsbrenn to turn the technology over to the
goverment and keep everything secret. They went nuts and ended up
defecting to the USSR. The FBI almost caught them but they made it to
Berlin. They had to leave their equipment behind though. Then, the
plane that was taking them from east Berlin to Moskow crashed. Their
research papers probably went up in smoke with them. Back in the
states the FBI refused to deliver the confiscated equipment to the
NSA. After months of squabbling the FBI admitted they had gotten the
cased mixed up with another and had 'lost' the evidence. The story was
in the major papers at the time, but the public was paying more
attention to other matters, like the Russians invasion of Afghanistan,
and most people had nothing to do with computers back then, so its no
wonder almost no one remembers nowadays. Unfortunately the technical
details are unknown, but one must guess that they had come up with
something really special since there seems to be no doubt that the
device really worked.
Phil Carmody
> [snip]
> > this could decompress to. Such further information is contained in the
> > filename, see the aforementioned BARF compressor for a working example.
>
> What are you talking about? The filename is part of the file. You
> cannot achieve compression that way. Look, BARF is a joke. I guess
> Matt Mahoney called it that way after the Baruff-Hogsbrenn device.
Barf means puke/vomit (verb, but usable as a noun in its non-finite form)
I'm guessing that's all Matt was alluding to.
Phil
--
Unpatched IE vulnerability: Extended HTML Form Attack
Description: Cross Site Scripting through non-HTTP ports, stealing
cookies, etc.
Published: February 6th 2002
Reference: http://eyeonsecurity.org/advisories/multple-web-browsers-vulnerable-to-extended-form-attack.htm
Matt Mahoney
"Phil Carmody" <thefatphi...@yahoo.co.uk> wrote in message
news:87ektct...@nonospaz.fatphil.org...
> guenther.v...@gmx.de (Guenther von Knakspott) writes:
>
> > [snip]
> > > this could decompress to. Such further information is contained in
the
> > > filename, see the aforementioned BARF compressor for a working
example.
> >
> > What are you talking about? The filename is part of the file. You
> > cannot achieve compression that way. Look, BARF is a joke. I guess
> > Matt Mahoney called it that way after the Baruff-Hogsbrenn device.
>
> Barf means puke/vomit (verb, but usable as a noun in its non-finite form)
> I'm guessing that's all Matt was alluding to.
It does? I thought it meant Better Archiver with Recursive Functionality.
But I did work it into an adjective. :-)
-- Matt Mahoney
Guenther von Knakspott
>
> > [snip]
> > > this could decompress to. Such further information is contained in the
> > > filename, see the aforementioned BARF compressor for a working example.
> >
> > What are you talking about? The filename is part of the file. You
> > cannot achieve compression that way. Look, BARF is a joke. I guess
> > Matt Mahoney called it that way after the Baruff-Hogsbrenn device.
>
> Barf means puke/vomit (verb, but usable as a noun in its non-finite form)
> I'm guessing that's all Matt was alluding to.
>
> Phil
But where is the relationship? Isn't there more affinity between an
algorithm and a device which actually performs that what the algorithm
mocks?, than between regurgitation and an algorithm?
Of course, we could wait and see if Matt Mahoney cares to clear up the
question. On the other hand the leitmotiv of this thread has rather
been setting up wild speculations...
Guenther von Knakspott
>
> > [snip]
> > > this could decompress to. Such further information is contained in the
> > > filename, see the aforementioned BARF compressor for a working example.
> >
> > What are you talking about? The filename is part of the file. You
> > cannot achieve compression that way. Look, BARF is a joke. I guess
> > Matt Mahoney called it that way after the Baruff-Hogsbrenn device.
>
> Barf means puke/vomit (verb, but usable as a noun in its non-finite form)
> I'm guessing that's all Matt was alluding to.
>
> Phil
But where is the relationship? Isn't there more affinity between an