Home > Archive > Compression > March 2007 > Don Knuth - Uncompressable sequences
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Don Knuth - Uncompressable sequences
|
|
| NickHolby@googlemail.com 2007-03-08, 6:56 pm |
| Hi,
I'm fairly new to the compression playground and I've ran into this:
http://www.wired.com/news/technolog...2,49599,00.html
about ZeoSyncs claim, which if their website is anything to go by now
fell through. ;) Although I'm more interested in this part:
"We can compress sequences that Stanford University professor Don
Knuth described as uncompressible (in his Art of Computer
Programming)."
What are the sequences that are classed as uncompressable? If this
infringes on copyright etc... then not too worry. :)
Thanks,
NH.
| |
| Josiah Carlson 2007-03-09, 9:55 pm |
| NickHolby@googlemail.com wrote:
> I'm fairly new to the compression playground and I've ran into this:
> http://www.wired.com/news/technolog...2,49599,00.html
>
> about ZeoSyncs claim, which if their website is anything to go by now
> fell through. ;) Although I'm more interested in this part:
>
> "We can compress sequences that Stanford University professor Don
> Knuth described as uncompressible (in his Art of Computer
> Programming)."
>
> What are the sequences that are classed as uncompressable? If this
> infringes on copyright etc... then not too worry. :)
One of the ways of modeling data is that symbols seen in the input are
generated randomly with a particular distribution. If the distribution
is biased (with or without context, as defined by Markov et al.),
entropy-based coders (Shannon-Fano, Huffman, Arithmetic) are able to
reduce their size.
Other ways of modeling data expand the idea of symbols and probabilities
to include words, dictionaries of words, etc. (LZ77, LZ78, LZW, etc.),
or start with a general data transformation to make the data "more
compressible" (BWT, bitwise-BWT, SZIP's partial BWT, etc.), or use
signal transformations to decompose the data into signals (DFT, DCT,
Wavelets, etc.) but typically are finished by an entropy coder stage.
Without having read that particular portion of Donald Knuth's volumes, I
would presume that he was talking about data that was random. That is,
the probability of seeing any symbol is equal, regardless of context,
regardless of what comes before, after, etc. Despite what certain
individuals would claim, compressing random data is (as per information
theory) impossible. You can come to it being impossible from different
angles, from a counting argument, to a Shannon information theory
0-order entropy calculation, etc., but the result is the same; making
*any* choice takes at least one bit. Even if you are choosing a
standard compression algorithm, or a specialized algorithm that
*expects* encrypted data, that takes a bit. For claimed recursive
compressors, it gets worse. Somewhere you need to encode how many
passes it took, which takes even more data.
- Josiah
| |
| Claudio Grondi 2007-03-11, 6:56 pm |
| NickHolby@googlemail.com wrote:
> Hi,
>
> I'm fairly new to the compression playground and I've ran into this:
> http://www.wired.com/news/technolog...2,49599,00.html
>
> about ZeoSyncs claim, which if their website is anything to go by now
> fell through. ;) Although I'm more interested in this part:
>
> "We can compress sequences that Stanford University professor Don
> Knuth described as uncompressible (in his Art of Computer
> Programming)."
>
> What are the sequences that are classed as uncompressable? If this
> infringes on copyright etc... then not too worry. :)
To classify data patterns as incompressible it is necessary to define
the compression methods allowed to be applied to the patterns.
To know what ZeoSyncs is speaking about, it is necessary to have read
Knuth's 'Art of Computer Programming' (sorry, I haven't).
As you are 'fairly new to the compression playground' it is maybe worth
to mention that :
The property of being compressible or incompressible is not a property
of a data pattern - it is a property of the _relationship_ of the data
pattern to the algorithms used for its compression.
Proper understanding of this above helps much to avoid endless and
fruitless discussions leading to nowhere - going through the past
traffic in this newsgroup will provide you with appropriate evidence and
hopefully allow already 'in advance' to see what is worth to talk about
and what is not.
Claudio Grondi
| |
| Phil Carmody 2007-03-11, 6:56 pm |
| NickHolby@googlemail.com writes:
> Hi,
>
> I'm fairly new to the compression playground and I've ran into this:
> http://www.wired.com/news/technolog...2,49599,00.html
>
> about ZeoSyncs claim, which if their website is anything to go by now
> fell through. ;) Although I'm more interested in this part:
>
> "We can compress sequences that Stanford University professor Don
> Knuth described as uncompressible (in his Art of Computer
> Programming)."
>
> What are the sequences that are classed as uncompressable? If this
> infringes on copyright etc... then not too worry. :)
Given that ZeoSync knowingly (to themselves, and to the cogniscenti)
lie through their teeth, my bet is that there are no such sequences,
unless they're deliberately misquoting the bible.
Given that there are over 2000 pages, for them to not provide a
reference to even the chapter and section is so sloppy I'm almost
inclined to think that they'd be lying their teeth again. Oh, I
was already.
Phil
--
"Home taping is killing big business profits. We left this side blank
so you can help." -- Dead Kennedys, written upon the B-side of tapes of
/In God We Trust, Inc./.
| |
| Thomas Richter 2007-03-11, 6:56 pm |
| NickHolby@googlemail.com wrote:
> I'm fairly new to the compression playground and I've ran into this:
> http://www.wired.com/news/technolog...2,49599,00.html
>
> about ZeoSyncs claim, which if their website is anything to go by now
> fell through. ;) Although I'm more interested in this part:
>
> "We can compress sequences that Stanford University professor Don
> Knuth described as uncompressible (in his Art of Computer
> Programming)."
>
> What are the sequences that are classed as uncompressable? If this
> infringes on copyright etc... then not too worry. :)
I don't remember Knuth said anything about "compressing" sequences (yes,
I did read the books). What this statement is relating to is most likely
the second volume of the work, namely the chapter about random
generators and random sequences. Knuth here discusses a couple of
heuristic tests to check the quality of a PseudoRNG, and what the above
statement means is likely that there are sequences that pass those tests
that can be still compressed. However, this isn't really surprising. It
isn't hard to generate PseudoRNGs that are easly compressible - namely
just by one number, the seed of the RNG and the code for the RNG. Works
perfectly. (-;
Thus, in other words: Sounds like "hot air" for me, and a statement that
is easily "compressible". (-:
So long,
Thomas
| |
| NickHolby@googlemail.com 2007-03-11, 6:56 pm |
| Ah okay. That's been much help, thanks everybody for taking time to
explain that. :) Since ZeoSync actually had applied for a patent, (and
I think it was granted?) excluding the perfect compression claim, did
any of the techniques outlined in the patent actually offer any new
insight for practical compression techniques?
NH.
| |
|
|
| Matt Mahoney 2007-03-22, 9:55 pm |
| On Mar 11, 3:15 pm, NickHo...@googlemail.com wrote:
> Ah okay. That's been much help, thanks everybody for taking time to
> explain that. :) Since ZeoSync actually had applied for a patent, (and
> I think it was granted?) excluding the perfect compression claim, did
> any of the techniques outlined in the patent actually offer any new
> insight for practical compression techniques?
>
> NH.
There are several patents on compression of random data. None of them
actually work.
http://www.faqs.org/faqs/compressio.../section-8.html
-- Matt Mahoney
| |
| tan_steve@yahoo.com 2007-03-23, 6:55 pm |
| On Mar 22, 2:31 pm, "Matt Mahoney" <matmaho...@yahoo.com> wrote:
> On Mar 11, 3:15 pm, NickHo...@googlemail.com wrote:
>
>
>
> There are several patents on compression of random data. None of them
> actually work.http://www.faqs.org/faqs/compressio.../section-8.html
>
> -- Matt Mahoney
Algorithmically, is it not possible that there exists an oracle, that
given a sha or other hash of the input sequence, can predict or
reconstruct the original sequence regardless of the length of the
original sequence?
| |
| Josiah Carlson 2007-03-23, 6:55 pm |
| tan_steve@yahoo.com wrote:
> Algorithmically, is it not possible that there exists an oracle, that
> given a sha or other hash of the input sequence, can predict or
> reconstruct the original sequence regardless of the length of the
> original sequence?
No. There only exists 2^160 sha1 hashes. Under the assumption that it
produces as few collisions as possible, you would only be able to
identify the 2^160 unique 20 byte files. Even if you specified the
length (and assuming that the sha1 hash was uniformly distributed), you
would have 256 possible matches for a 21 byte file with a particular
sha1 hash, 65536 for a 22 byte file, etc.
Despite what the internet says, Bruce Schneier does not see sha1 as an
invertible compression algorithm. :)
- Josiah
| |
| Mark Adler 2007-03-23, 6:55 pm |
| On Mar 8, 9:58 am, NickHo...@googlemail.com wrote:
> "We can compress sequences that Stanford University professor Don
> Knuth described as uncompressible (in his Art of Computer
> Programming)."
>
> What are the sequences that are classed as uncompressable?
The press release (which has subsequently disappeared, along with the
entire zeosync web site) referenced the 1973 edition of volume 3,
Sorting and Searching, pp. 506-542.
I have the 1997 edition, and it's not obvious to me where there would
be any such mention. If anyone has the 1973 edition, maybe they can
check those pages. I don't recall a mention of uncompressable
sequences, certainly not in Volume 3. (Volume 2 is the one that deals
with random sequences).
Mark
| |
|
|
| markn@ieee.org 2007-03-24, 9:56 pm |
| On Mar 23, 4:50 pm, "Mark Adler" <mad...@alumni.caltech.edu> wrote:
> The press release (which has subsequently disappeared, along with the
> entire zeosync web site) referenced the 1973 edition of volume 3,
> Sorting and Searching, pp. 506-542.
>
> I have the 1997 edition, and it's not obvious to me where there would
> be any such mention. If anyone has the 1973 edition, maybe they can
> check those pages. I don't recall a mention of uncompressable
> sequences, certainly not in Volume 3. (Volume 2 is the one that deals
> with random sequences).
I have the 1973 edition, the chapter on Hashing runs from p. 506 to p.
542.
Naturally there are many mentions of the word "random" but no
discussion of compressibility that I can find. (nor in the exercises.)
I did learn one interesting historical fact: the word "hashing" had
apparently never seen print until Robert Morris (the elder) published
his survey paper in CACM 1968 on the topic. Knuth seems to imply that
up until that time it was in common use, but was seen as slang or
jargon, and thus not particularly dignified.
|
| Mark Nelson - http://marknelson.us
|
| |
| byaarov@yahoo.com 2007-03-28, 3:56 am |
| On Mar 23, 8:46 am, Josiah Carlson <josiah.carl...@sbcglobal.net>
wrote:
> tan_st...@yahoo.com wrote:
>
> No. There only exists 2^160 sha1 hashes. Under the assumption that it
> produces as few collisions as possible, you would only be able to
> identify the 2^160 unique 20 byte files. Even if you specified the
> length (and assuming that the sha1 hash was uniformly distributed), you
> would have 256 possible matches for a 21 byte file with a particular
> sha1 hash, 65536 for a 22 byte file, etc.
>
> Despite what the internet says, Bruce Schneier does not see sha1 as an
> invertible compression algorithm. :)
>
> - Josiah
This makes me wonder... if the start point of the universe were
treated as a compressed string, then somehow there exists some
compression algorithm the universe is using that is causing that small
piece of information to expand constantly, into something that
obviously isn't just noise. While I understand the math behind why an
oracle could not magically map small numbers to all possible
combinations of strings greater than what the hash can represent, I
still must admit that I am  by how, if evolution and the
theory of the universe is true, how so much expansion has been done
from possibly just one bit of information in the beginning.
Unless the universe were a program as big as itself. In which case,
who created that program and how big were they?
Can all this must mean only one thing... there must be a god, and the
theories of the universe are wrong?
I think I need to lay off the bong :)
| |
| Josiah Carlson 2007-03-28, 3:56 am |
| byaarov@yahoo.com wrote:
> This makes me wonder... if the start point of the universe were
> treated as a compressed string, then somehow there exists some
> compression algorithm the universe is using that is causing that small
> piece of information to expand constantly, into something that
> obviously isn't just noise. While I understand the math behind why an
> oracle could not magically map small numbers to all possible
> combinations of strings greater than what the hash can represent, I
> still must admit that I am by how, if evolution and the
> theory of the universe is true, how so much expansion has been done
> from possibly just one bit of information in the beginning.
But it wasn't just one bit. Or a thousand bits. In order to correctly
characterize the universe at t = 0 (the point of the big bang), one
would need to describe the position, orientation, velocity, bonds, etc.,
of all elementary particles (protons, neutrons, electrons, photons, or
perhaps quarks and leptons, etc.) *exactly*. Where "exactly" is perhaps
limited to 1/1000 the diameter of the smallest quark (if that makes
sense). But there are at least 1.5*10**79 atoms in the observable
universe, never mind photons, em fields, gravity fields, etc., whose
state is a function of the location of the above plus the fields,
moments ago.
I also don't believe that the universe has gained more information, it
has just become more ordered due to certain intrinsic continuous
physical laws and element/mollecular properties (densities, melting
points, etc.), and the "information" that is being "created" is really
just the redirection of energy to organize information in another way.
As for evolution, we can look at bacterial evolution, pet breeding, or
any one of the thousands of examples of obvious genetic adaptation to
environments to show that it happens all the time.
> Unless the universe were a program as big as itself. In which case,
> who created that program and how big were they?
>
> Can all this must mean only one thing... there must be a god, and the
> theories of the universe are wrong?
Why does there need to be a creator? And why must the existence of a
creator imply that scientific theories are wrong? There doesn't, and it
doesn't. Certain things become more convenient with the existence of a
creator, but many more questions arise. Also, the existence of a
creator doesn't imply the existence of heaven, hell, reincarnation, etc.
(any of would also be quite convenient), as we may be seen as little
more than bacteria to a creator.
> I think I need to lay off the bong :)
That sounds like a good idea.
- Josiah
| |
| byaarov@yahoo.com 2007-03-28, 3:56 am |
| On Mar 27, 4:39 pm, Josiah Carlson <josiah.carl...@sbcglobal.net>
wrote:
> byaa...@yahoo.com wrote:
>
> But it wasn't just one bit. Or a thousand bits. In order to correctly
> characterize the universe at t = 0 (the point of the big bang), one
> would need to describe the position, orientation, velocity, bonds, etc.,
> of all elementary particles (protons, neutrons, electrons, photons, or
> perhaps quarks and leptons, etc.) *exactly*. Where "exactly" is perhaps
> limited to 1/1000 the diameter of the smallest quark (if that makes
> sense).
Is it not useful to ask what piece of information created those
particles and the force for the velocity?
> But there are at least 1.5*10**79 atoms in the observable
> universe, never mind photons, em fields, gravity fields, etc., whose
> state is a function of the location of the above plus the fields,
> moments ago.
>
> I also don't believe that the universe has gained more information, it
> has just become more ordered due to certain intrinsic continuous
> physical laws and element/mollecular properties (densities, melting
> points, etc.), and the "information" that is being "created" is really
> just the redirection of energy to organize information in another way.
>
> As for evolution, we can look at bacterial evolution, pet breeding, or
> any one of the thousands of examples of obvious genetic adaptation to
> environments to show that it happens all the time.
>
>
>
> Why does there need to be a creator? And why must the existence of a
> creator imply that scientific theories are wrong? There doesn't, and it
> doesn't. Certain things become more convenient with the existence of a
> creator, but many more questions arise. Also, the existence of a
> creator doesn't imply the existence of heaven, hell, reincarnation, etc.
> (any of would also be quite convenient), as we may be seen as little
> more than bacteria to a creator.
>
>
> That sounds like a good idea.
>
> - Josiah
| |
| Josiah Carlson 2007-03-28, 3:56 am |
| byaarov@yahoo.com wrote:
> On Mar 27, 4:39 pm, Josiah Carlson <josiah.carl...@sbcglobal.net>
> wrote:
>
> Is it not useful to ask what piece of information created those
> particles and the force for the velocity?
Unfortunately we do not have reasonable approximations as to the
probability of whether there existed a creator (information, god, or
otherwise), so we cannot use Shannon information theory to determine
whether the information content (using Shannon information theory :) of
such is significant.
Personally, I'll stick with it being inconsequential and/or dangerous.
Why? Because knowing whether or not there existed a creator would only
embolden those portions of humanity who claim that they know the one and
only truth, and s to subjugate and/or kill to "prove" it (or prove
that a non-existence is wrong).
The actual useful information are the thousands of questions that arise
after the first question is answered. But like the first question, even
if one has the true answer, someone is willing to kill or die to prove
or disprove it. I believe that it would be better to not know the
actual answer - it keeps some people busy asking the questions and not
killing each other. Then again, some people don't think enough to ask
the questions, and kill anyways.
- Josiah
| |
|
| Who has created the creator?
Cheers,
| |
| Willem 2007-03-28, 6:56 pm |
| CBS wrote:
) Who has created the creator?
We (mankind) did. Or rather, we *will*, in the far future. And this
Creator that we will have created, will not be bound by time and causality,
and will thus, inevitably being who s/h/it is, create the universe.
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
| |
| guenther vonKnakspott 2007-03-28, 6:56 pm |
| On Mar 28, 9:33 am, Josiah Carlson <josiah.carl...@sbcglobal.net>
wrote:
> byaa...@yahoo.com wrote:
>
>
> Unfortunately we do not have reasonable approximations as to the
> probability of whether there existed a creator (information, god, or
> otherwise), so we cannot use Shannon information theory to determine
> whether the information content (using Shannon information theory :) of
> such is significant.
>
> Personally, I'll stick with it being inconsequential and/or dangerous.
> Why? Because knowing whether or not there existed a creator would only
> embolden those portions of humanity who claim that they know the one and
> only truth, and s to subjugate and/or kill to "prove" it (or prove
> that a non-existence is wrong).
> The actual useful information are the thousands of questions that arise
> after the first question is answered. But like the first question, even
> if one has the true answer, someone is willing to kill or die to prove
> or disprove it. I believe that it would be better to not know the
> actual answer - it keeps some people busy asking the questions and not
> killing each other. Then again, some people don't think enough to ask
> the questions, and kill anyways.
>
> - Josiah
You seem to confuse knowledge in the general sense with knowledge
acquired by science. Some of us know that God exists because we
believe. Others know that God doesn't exist because they have
determined this fact through their philosophy. Science is a particular
bad way of approaching the question if you ask me. But in all these
three camps and many others, you will find those who try to coerce
others into accepting their knowledge, and those who will not. It is a
human trait that has nothing to do with knowledge of a particular
sort. And it is also not being kept busy that keeps some people from
coercing others, but the fact that people who ask questions tend to be
more inclined towards learning than towards forcing others into
acceptance. Or do you really think there are people around who would
rather be gutting infidels if they had any time to spare from their
studies?
Regards.
| |
| Phil Carmody 2007-03-28, 6:56 pm |
| Willem <willem@stack.nl> writes:
> CBS wrote:
> ) Who has created the creator?
>
> We (mankind) did. Or rather, we *will*, in the far future. And this
> Creator that we will have created, will not be bound by time and causality,
> and will thus, inevitably being who s/h/it is, create the universe.
Excellent! Will this creator be able to violate the Kraft inequality?
If so, I foresee some exciting times ahead for the field of compression,
and in particular this newsfroup!
Phil
--
"Home taping is killing big business profits. We left this side blank
so you can help." -- Dead Kennedys, written upon the B-side of tapes of
/In God We Trust, Inc./.
| |
|
| Josiah Carlson ha scritto:
> byaarov@yahoo.com wrote:
>
> I also don't believe that the universe has gained more information, it
> has just become more ordered due to certain intrinsic continuous
> physical laws and element/mollecular properties (densities, melting
> points, etc.), and the "information" that is being "created" is really
> just the redirection of energy to organize information in another way.
>
>
Making order from chaos *does* require information. Better, maybe that's
the only information needed at all.
Being this information at the beginning or provided from time to time,
it has to be there, in the process of "making order from chaos".
Einstein rightly said: "It doesn't matter to know the thing
(information) itself, but where to find it" (not an exact quote)
Best,
E.
| |
| byaarov@yahoo.com 2007-03-28, 6:56 pm |
| On Mar 28, 8:08 am, erpy <i...@forwardgames.com> wrote:
> Josiah Carlson ha scritto:
>
>
>
> Making order from chaos *does* require information. Better, maybe that's
> the only information needed at all.
> Being this information at the beginning or provided from time to time,
> it has to be there, in the process of "making order from chaos".
>
> Einstein rightly said: "It doesn't matter to know the thing
> (information) itself, but where to find it" (not an exact quote)
>
> Best,
> E.
I guess all I am wondering is this ... If you treat the ever expanding
universe as a form of decompression, then at time t(0) if the amount
of information was x and the expansion program were y, then you are
saying that the amount of information contained with x + y is exactly
the same amount of information contained in the universe today. Then
two things dont make sense here... one, where did that combination of
information itself come from? That is, what program created x or y?
Or could the program y that created the universe U from x have been a
complete random act. Does that mean that random() is a form of
decompression?
| |
|
| byaarov@yahoo.com ha scritto:
> On Mar 28, 8:08 am, erpy <i...@forwardgames.com> wrote:
>
>
>
> I guess all I am wondering is this ... If you treat the ever expanding
> universe as a form of decompression, then at time t(0) if the amount
> of information was x and the expansion program were y, then you are
> saying that the amount of information contained with x + y is exactly
> the same amount of information contained in the universe today. Then
> two things dont make sense here... one, where did that combination of
> information itself come from? That is, what program created x or y?
>
> Or could the program y that created the universe U from x have been a
> complete random act. Does that mean that random() is a form of
> decompression?
>
>
All I said is, that it doesn't matter "what" you store at the beginning
in terms of "number of items".
There could have been a bunch of elementary particles. It's the rules
that mattered.
It doesn't have to be "random". You can create any algorithm that
generates an infinite ordered piece of information...you just let it run.
The "geniality" here is in formulating the rules, so that the so
generated information contains both "some order" and "pseudo random"
behaviors that might eventually get ordered by the rules.
In terms of compression, this doesn't mean that the initial rules are a
"universal" compressor....just because they can "only" describe the
whole system as it is instant by instant, and not all the possible
(not-happened) variations of it. (like a generic compressor *tries* to
do on the digital side)
It's like saying that the "rules" can generate just 1 particular file,
of all possible files, even if infinitely long - thus getting the best
compression ratio. :)
My opinion is that "physics" still misses most part of these rules, or
the "global meaning" of those, thus it cannot properly study the nature
of this "algorithm".
Best,
E.
| |
| Josiah Carlson 2007-03-30, 3:55 am |
| guenther vonKnakspott wrote:
[snip]
> You seem to confuse knowledge in the general sense with knowledge
> acquired by science.
I do not.
Not all the knowledge that I possess was acquired by science, but it can
all be readily divided into experience and heresay. Anything that I
have not experienced myself, I consider to be heresay. But there are
varying levels of the quality of the heresay, usually a function of
trust and testability.
If it is in a calculus textbook on my shelf, that's pretty good heresay,
even if I can't directly touch the infinitesimals. Or if my wife tells
me that there's some mold in a bowl in the fridge. But if its a report
on the internet that a 12 year old boy has managed to build a rocket
ship and travel to the moon, I'm going to need some harder evidence to
believe.
> Some of us know that God exists because we
> believe. Others know that God doesn't exist because they have
> determined this fact through their philosophy.
I think you are squeezing people into categories that are too narrow.
Sometimes it isn't about knowing or not knowing, it's about belief,
faith, and/or experience. (also coercion)
> Science is a particular
> bad way of approaching the question if you ask me.
Of course science is a poor way of approaching the question. Science is
about hypothesis, theory, and verification or disproving via testing.
The usual way of defining god is that it is not testable with scientific
instruments or able to be experienced by our senses. It's not that our
scientific tools aren't sufficient, it's that the definition of god is
such that it is not testable. This isn't a fault with science, or the
tools used by scientists (despite claims to the contrary by certain
people, though none that I've seen post here).
Is it the fault of religion? That all depends on the purpose of
religion. If religion isn't supposed to be provable, then an untestable
god is fine. But if religion is supposed to be provable, then the
definition of god as untestable is a serious problem.
> But in all these
> three camps and many others, you will find those who try to coerce
> others into accepting their knowledge, and those who will not. It is a
> human trait that has nothing to do with knowledge of a particular
> sort.
People like to talk about what they know. Whether that talking (or
teaching) involves coercion depends on the intent of both parties. If
your teacher is trying to teach you calculus, and you don't want to
learn it, the teacher is probably going to try to coerce you. I doubt
you will find any reasonable person who would find this coercion bad.
But when it comes to teaching anything involving religion (in favor of
one, in favor of questioning religion, etc.), depending on the intent of
the "student", could it reasonably be considered teaching? Coercion?
Something in between? Does age matter? Does maturity matter?
While many westerners have been looking at the middle east and the
jihadists, wondering how to handle the situation, the US has the exact
same problem with certain groups of fundamental christians - who are
_coercing_ their children to live and die (its only a few more steps to
kill) for the word of their god (really the word of parents, priests, etc.).
> And it is also not being kept busy that keeps some people from
> coercing others, but the fact that people who ask questions tend to be
> more inclined towards learning than towards forcing others into
> acceptance.
Indeed. Keep asking questions. Keep asking what would happen if you
were to know the answer. Who knows, you may even be able to come up
with a convincing enough answer (be it true or not ;).
> Or do you really think there are people around who would
> rather be gutting infidels if they had any time to spare from their
> studies?
While some would much rather study than kill, there are those who
happily kill others because of what they believe, religious or not. As
an example, take your pick of any ethnic slaughter, cru e, or war.
On the one hand I can understand, I would defend those I love. But when
someone claims to love their god and country, and that they must defend
their country from the infidels, blasphemers, etc. (even when they are
half a world away), a reasonable person has to question this would-be
'defender of all' (which is applicable to both sides of the current "war
on terror").
Maybe I've been researching too long, or maybe I've become lazy, or
maybe I'm too much of a pragmatist, but I find that I prefer to dig into
questions that I have a chance of solving. The question of whether
there existed a creator, or whether that creator can be known, or the
thousands of questions that come from knowing the answers to those,
really have no provable answer. As a personal choice, I tend to prefer
questions with answers that I can can solve, and whose solutions I can
prove correct. Not for others (though that is a convenient secondary
benefit), but for myself.
- Josiah
| |
| Claudio Grondi 2007-03-30, 3:55 am |
| byaarov@yahoo.com wrote:
> On Mar 28, 8:08 am, erpy <i...@forwardgames.com> wrote:
>
>
> I guess all I am wondering is this ... If you treat the ever expanding
> universe as a form of decompression, then at time t(0) if the amount
> of information was x and the expansion program were y, then you are
> saying that the amount of information contained with x + y is exactly
> the same amount of information contained in the universe today. Then
> two things dont make sense here... one, where did that combination of
> information itself come from? That is, what program created x or y?
From my point of view a much more intriguing question in this context
is, where comes your desire to get an answer to your last question from?
What will you consider as gain if actually get somehow to know the answer?
>
> Or could the program y that created the universe U from x have been a
> complete random act. Does that mean that random() is a form of
> decompression?
I can't clearly see what the second statement has to do with the first
one, but yes, random() is a form of decompression - it takes the
information where to start 'In-1'(implicit or explicit) and the
requested size of the output 'In-2' (implicit or explicit), so naming
the output data 'Out', random() is nothing else a a table with three
columns:
'In-1' | 'In-2' | 'Out' |
in-1-1 in-2-1 out-1
in-1-2 in-2-2 out-2
in-1-3 in-2-3 out-3
... ... ...
where the input
in-1-# and in-2-#
has fixed size and where in case of large values for in-2-# the size of
out-# is much larger than the size of in-1-# and in-2-# together.
To give you another way of considering what random() is imagine, that it
is nothing else as a very large chunk of data (size depends here on how
random() is actually programmed and after which size of output it starts
to repeat the same data pattern). From this point of view in-1-# is an
address of the initial datum in that chunk of data which is written to
out-# and in-2-# is the size of the data written to out-# beginning at
in-1-#.
The only information you need to generate that very large chunk of data
for usage as described above is the algorithm behind random(). From this
point of view you can consider random() the most effective compression
scheme among all other ones for storing data patterns being sub-patterns
of the data pattern within the very large chunk of data it is able to
generate.
Hope this helps to gain some better understanding of this random()
thingy. Does it?
Claudio
| |
| Claudio Grondi 2007-03-30, 3:55 am |
| Josiah Carlson wrote:[color=darkred]
> guenther vonKnakspott wrote:
> [snip]
>
> I do not.
>
> Not all the knowledge that I possess was acquired by science, but it can
> all be readily divided into experience and heresay. Anything that I
> have not experienced myself, I consider to be heresay. But there are
> varying levels of the quality of the heresay, usually a function of
> trust and testability.
Josiah,
reading your entire reply, I have to admit, that I have trouble to see
in your posting the added value when compared to the original message by
Günter von Knakspott and it appears to me somehow, that you still seem
not to get some of the by him so brilliantly formulated insights.
Just trying to guess why it appears to me, that you are still missing
some very basic understanding it comes to my mind, that you seem to
consider your own experience as something you can fully trust and in
this connection forget to consider, that even if it is your very private
own personal experience it is also built upon an unprovable belief
system you are maybe even not aware of.
---
Sorry if this above starts to be fully out of topic here, but
'uncompressable sequences' is a kind of subject that seems inevitably
lead to widespread side effects.
Claudio Grondi
| |
| Josiah Carlson 2007-03-30, 9:55 pm |
| Claudio Grondi wrote:
> Josiah Carlson wrote:
>
> Josiah,
> reading your entire reply, I have to admit, that I have trouble to see
> in your posting the added value when compared to the original message by
> Günter von Knakspott and it appears to me somehow, that you still seem
> not to get some of the by him so brilliantly formulated insights.
Being that I'm a not a philosopher (I'm a computer scientist), missing
some or all of Gunter's brilliantly formulated insights wouldn't be a
surprise. Though it seems to me that the first part was a claim that I
don't acknowledge non-experiential knowledge (I do), followed by a brief
false dichotomy of science vs. religion (the dichotomy really only
exists for people who want it to exist), two sentences about teaching
and knowledge passing as being implicitly coercive, followed by
statements claiming that I believe people would rather kill than study.
What I should have said prior to the message that Gunter posted, which I
attempted to state afterwards, is that whether or not one asks questions
about god, one must acknowledge that the answers are tacitly unprovable
(by definitions and presuppositions about the nature of god). I would
encourage everyone to ask questions about their god and their faith, be
they christian, jewish, muslim, ba'hai, hindu, athiest, or otherwise.
But its not a matter of questions or faith, its that so many people
claim to have the answers to these questions, that actually knowing the
first answer (who or what is the creator) doesn't get you very far, and
when you try to share the "one true answer" that you have found, you can
easily find yourself in trouble.
But if you want to stick with "Josiah isn't as concise, elegant, or
insightful as Gunter", that's fine too. I've been called worse than
verbose, inelegant and not insightful. I would highly suggest calling
me "insufferable prick", it is both short and likely precise.
> Just trying to guess why it appears to me, that you are still missing
> some very basic understanding it comes to my mind, that you seem to
> consider your own experience as something you can fully trust and in
> this connection forget to consider, that even if it is your very private
> own personal experience it is also built upon an unprovable belief
> system you are maybe even not aware of.
Maybe I'm missing something (this wouldn't be the first time), but it
seems as if we both agree that all perspectives are based on certain
presuppositions (I seem to have failed to mention this earlier). I
presuppose that my experiences, senses, and my interpretations of such
produce a reasonable facsimile of truth (or something close enough for
me to get by). Perhaps others presuppose that some book or clergy and
their interpretations of it offers the same.
You seem to be saying that I don't acknowledge my presuppositions. I
didn't mention my presuppositions because I honestly didn't think that I
would have to explain that I trusted my experience - I assumed that most
people trust a majority of their experiences, as going through life not
trusting experience would be a bit like not knowing if your next step
will land on the ground or in the sky.
> Sorry if this above starts to be fully out of topic here, but
> 'uncompressable sequences' is a kind of subject that seems inevitably
> lead to widespread side effects.
It's only as off-topic as most of the other posts in this thread.
Speaking of which, I'm probably going to bow out unless someone talks
about data compression again, though not before posting a link to an
interview with one of the lead scientists on the human genome project,
who doesn't seem to have an issue reconciling his scientific experience
with his faith (and who speaks more elegantly than I):
http://www.npr.org/templates/story/...storyId=9207913
I hope he doesn't need to state his presuppositions that he trusts his
experiences ;)
Good day,
- Josiah
|
|
|
|
|