Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Illegal Numbers

en.wikipedia.org

Post reply on HN
Print view View original

61–70 of 98 posts

Re: Illegal Numbers

#61

Post by lmm »

Evbn wrote: post #31
rolux wrote: post #17

The main problem with "illegal numbers" seems to be that for any illegal number a , which represents someone's intellectual property or trade secret A , I can easily create two original pieces of intellectual property B and C so that b + c = a .

Give us an actual example of b and c that are created without knowledge of a, and you may have case. But information theory says it is unlikely, if a is deeply original and not derivative of something like b and c to begin with.

I've read an argument that the law has considered a five(?)-note sequence enough to identify one song as a derivative of another - and given reasonable estimates for the "space" of possible 3-minute tunes, and the number of (copyrighted) songs currently in existence, there's a significant chance that a "random" new song will be legally considered a derivative of some existing song.

Of course, here we come back to the "color" of bits; one is unlikely to be prosecuted for "copying" a song one has never heard. Though cases like My Sweet Lord are enough to give one pause.

Re: Illegal Numbers

#62

Post by aes256 »

wmf wrote: post #46
aes256 wrote: post #41

Earlier quoted context omitted.

Apparently what I was describing here has already been put into practice in the shape of Monolith [1] In the case I describe, as in the case of 'munging' using Monolith, neither B nor C bears any resemblance to A except when the two are combined. In my view, at least, you cannot say either is a derivative of A. To do so would bind you to declare everything a derivative of A, because any B when combined with an approp…

Let me put it a different way. If A = decompress(B), then necessarily B = compress(A), so B is obviously a derivative of A. Introducing xor does not change anything; one of the parts must be a derivative of the original.

Okay, let me propose an alternative procedure.

I set a series of random number generators going, and with each set of results, I apply randomly generated XOR to create a new sequence of numbers.

I perform this process over and over. Eventually, it produces (give or take a few bits) a copy of an MP3 file of a copyrighted work.

Now, once we've eliminated any procedure of creating B and C that includes A, would you still say one of B or C are derivative of A?

Should my random number generator be banned? Perhaps more importantly, do I acquire copyright to all the files it creates?

I could quite easily create every possible variation of an MP3 file of a given length. Does that mean any musician who, using a different procedure, produces one of these files is infringing on my copyright?

Re: Illegal Numbers

#63

Post by aes256 »

Leszek wrote: post #54
aes256 wrote: post #33

Earlier quoted context omitted.

I've never really understood the child pornography thing. Criminalizing mere possession of such images seems like a roundabout, largely ineffective, and perhaps even counterproductive, way of tackling the real issue; the sexual abuse of children. It blurs the very real and very important distinction between those who are sexually attracted to children and those who will act on that attraction to sexually abuse a chil…

Fine, replace child pornography with a text file containing your name, birthday, social security number, credit card details, email password and mother's maiden name, the argument is the same.

As lmm says, I don't have any issue with someone possessing that info. Security through obscurity is folly, and it is only as a result of incompetent security procedures that someone could cause harm with these pieces of information.

Re: Illegal Numbers

#64

Post by vidarh »

iopq wrote: post #53
cletus wrote: post #13

This is an interesting problem that we, as a society, need to deal with. HN's readership would be predominantly from a tech background with probably a reasonable basis in mathematics and the sciences. As such we can recognize some of the absurdities of the legal system, like how you can't patent a mathematical formula but you can patent software, which is ultimately indistinguishable from a formula in many (if not al…

>So for a small integer to encode, say, an obscene image or a program that circumvents copy protection, it would require another program that actually does that. so winzip is the problem, not the zipped file? ridiculous statement, bro fine, you still need to play the file, so media player is the problem? obviously, it's the source file, not the generic tools that work on everything

He is saying that the problem is those components of the total system used to produce the infringing output that is not generic.

So in the case of a zipped file, the zipped file and not winzip would be the problem, because winzip is generic, and have general purpose uses, while the zipped file is a very large specific number that is translatable by a generic process into a specific protected work.

On the other extreme, the number "7" on the other hand, does not encode enough information to on its own contain the information about the infringing content. Even if you were to write a "decoder" that turns the number "7" into a movie, in this case the "decoder" is the problem, as the decoder embodies enough specific information about the content that in effect the number you feed in is little more than a key to unlock access to that content.

Re: Illegal Numbers

#65

Post by aes256 »

snowwrestler wrote: post #43
aes256 wrote: post #33

Earlier quoted context omitted.

I've never really understood the child pornography thing. Criminalizing mere possession of such images seems like a roundabout, largely ineffective, and perhaps even counterproductive, way of tackling the real issue; the sexual abuse of children. It blurs the very real and very important distinction between those who are sexually attracted to children and those who will act on that attraction to sexually abuse a chil…

Criminalizing the possession of child porn photos is a tool to assist in the capture of people who make child porn photos. A person facing prosecution for possession is highly motivated to reduce their own sentence by helping in the arrest and prosecution of the source of their images. If the source didn't make the photos, they certainly possessed them too, so it's possible to walk another link higher in the chain.

"You haven't caused any harm or done anything wrong, but we think you might have information that will lead us to someone who has, so we're going to invade your privacy, drag you through the courts, lock you up, and ultimately ruin your life."

I can't be the only one who sees something wrong there? The sexual abuse of children is horrible, but I don't it justifies the ritualistic destruction of countless adults' lives.

Re: Illegal Numbers

#66

Post by chmodd »

cletus wrote: post #13

This is an interesting problem that we, as a society, need to deal with. HN's readership would be predominantly from a tech background with probably a reasonable basis in mathematics and the sciences. As such we can recognize some of the absurdities of the legal system, like how you can't patent a mathematical formula but you can patent software, which is ultimately indistinguishable from a formula in many (if not al…

> I've had a discussion with someone else about how a group of people could route each other's packets and then hypothetically law enforcement couldn't prove who they came from.

Combine it with encryption and you'll get something like Tor or i2p. Not just hypothetical, but works very well in practice too. Things like Silk Road have been running there for years with LEOs not being able to do anything about it.

Re: Illegal Numbers

#67

Post by randomdata »

Evbn wrote: post #31
rolux wrote: post #17

The main problem with "illegal numbers" seems to be that for any illegal number a , which represents someone's intellectual property or trade secret A , I can easily create two original pieces of intellectual property B and C so that b + c = a .

Give us an actual example of b and c that are created without knowledge of a, and you may have case. But information theory says it is unlikely, if a is deeply original and not derivative of something like b and c to begin with.

If (b) and (c) can be considered derivative works, then should a low-bit hash of (a) not also be considered a derivative? If a given hashing algorithm says the hash of Metallica's latest hit is 42, have I just broken the law by posting this?

Re: Illegal Numbers

#68

Post by randomdata »

aes256 wrote: post #33
Leszek wrote: post #28

Let's rephrase this problem as not something that people will defend, e.g. anti-DRM, but something people will not defend; let's go straight for the big guns and use child pornography. Clearly, child pornography is illegal, and yet the images are just a series of 0s and 1s; these could potentially be re-interpreted as a single number, giving rise to the problem here. Or maybe each triplet of bytes of the JPEG's data…

I've never really understood the child pornography thing. Criminalizing mere possession of such images seems like a roundabout, largely ineffective, and perhaps even counterproductive, way of tackling the real issue; the sexual abuse of children. It blurs the very real and very important distinction between those who are sexually attracted to children and those who will act on that attraction to sexually abuse a chil…

The idea, however misguided, is that the consumption of child pornography creates a market demand. If nobody wants the images, nobody will be compelled to create them in the first place thereby eliminating the actual abuse.

Re: Illegal Numbers

#69

Post by nitrogen »

shasta wrote: post #36
JoshTriplett wrote: post #11

Earlier quoted context omitted.

From the perspective of the law, bits do have color, even though that makes no sense from a technical perspective. So, any given number may or may not qualify as "illegal" depending on how you generated it. That said, I don't think that makes the terms "illegal number" or "illegal prime" nonsensical, depending on context. In any case, I think the article doesn't deserve the pithy "summary" upthread; the summary provi…

My summary was snarky, but "illegal number" is a silly term intended to emphasize that "it's just a number (so how could it be illegal?)", which is as sophisticated as the legal arguments "it's just a plastic disk with certain reflective properties" or "it's just some markings on a piece of paper." Your claim that bits having color "makes no sense from a technical perspective" sounds like more geek misunderstanding t…

Your claim that bits having color "makes no sense from a technical perspective" sounds like more geek misunderstanding to me. There is no technical problem here -- only a problem of geeks applying technical results where they shouldn't.

How dare you suggest that geeks and their rationality are not the rightful masters of the universe?

Re: Illegal Numbers

#70

Post by nitrogen »

shasta wrote: post #50
JoshTriplett wrote: post #45

Earlier quoted context omitted.

> Since the argument here is over how the law should work I think we've talked past each other here. I agree with you about how the law works. As much as I might want it to work differently, it currently does not. However, read the second to last paragraph of the essay: "I think it's time for computer people to take Colour more seriously - if only so that we can better explain to the lawyers why they must give up the…

I think the few bits about lawyers not understanding the limitations of metadata weren't the thrust of the article. I do agree that the two viewpoints there aren't opposed, but that point about metadata is not the same as the wrong headed one being made by people using the phrase "illegal number."

The only thing that makes the "illegal number" view wrong headed is a belief that the law as currently on the books has the final say on what is ethically/factually/morally/intellectually right and wrong.

I believe you are too hard on those who disagree with you, as you don't seem to acknowledge that those who disagree with the law's desire to stamp a color on everything have valid reasons for their disagreement.

Post reply on HN