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

Illegal Prime Numbers

en.wikipedia.org

Post reply on HN
Print view View original

81–90 of 109 posts

Re: Illegal Prime Numbers

#81

Post by lostcolony »

privacylawthrow wrote: post #49
jasode wrote: post #17

There are several past threads of this article but this comment about how the DeCSS illegal prime was created seemed the most interesting: https://news.ycombinator.com/item?id=1045394 This might be thought of as a human version of a covert channel : https://en.wikipedia.org/wiki/Covert_channel E.g. publicizing the number and expiration date of Bill Gates' credit-card is illegal but what if we wrote a text commentary…

The law cares about intent and outcome. If you intended to publish Bill Gates' credit card number and did so, and if doing so was a crime, you'd be guilty of that crime. This is true regardless of how you published the information. There is no "out" for putting it behind a pretext. It's why sharing child pornography is illegal, even though all the creators are really doing is sharing a set of instructions for someone…

Sometimes the law cares about intent.

For copyright, it really doesn't. Hence Disney threatening to sue daycares for having Disney character murals on their walls.

While we have since had the four factors of fair use introduced, it still doesn't care about intent; it cares about how the work is used/transformed, but that's different than intent.

Re: Illegal Prime Numbers

#82

Post by snappieT »

lostcolony wrote: post #81
privacylawthrow wrote: post #49

Earlier quoted context omitted.

The law cares about intent and outcome. If you intended to publish Bill Gates' credit card number and did so, and if doing so was a crime, you'd be guilty of that crime. This is true regardless of how you published the information. There is no "out" for putting it behind a pretext. It's why sharing child pornography is illegal, even though all the creators are really doing is sharing a set of instructions for someone…

Sometimes the law cares about intent. For copyright, it really doesn't. Hence Disney threatening to sue daycares for having Disney character murals on their walls. While we have since had the four factors of fair use introduced, it still doesn't care about intent; it cares about how the work is used/transformed, but that's different than intent.

If you take copyrighted material and put it on a daycare mural, then you _intended_ to put copyrighted material on a daycare mural. It is not about intent to damage, it's just pure intent.

Re: Illegal Prime Numbers

#83

Post by bostonpete »

Zircom wrote: post #52
bostonpete wrote: post #45

Earlier quoted context omitted.

Just to be clear -- b/c I got a bit confused reading this -- this example seems to assume "12345" and "71" are base-256 numbers and "256" is base-10.

Once you compress the number it becomes stored in base-256 due to how the gzip file format works was my understanding

Sure, the benefit of base-256 is that a 0 maps to a byte with all 0's in the binary representation of the integer -- it's nothing specific to gzip. But on its face, 12345x256^8+71 = 1234500000071 is just wrong and only works once you realize that it's combining integers from different number systems.

Re: Illegal Prime Numbers

#84

Post by redwall_hp »

snappieT wrote: post #82
lostcolony wrote: post #81

Earlier quoted context omitted.

Sometimes the law cares about intent. For copyright, it really doesn't. Hence Disney threatening to sue daycares for having Disney character murals on their walls. While we have since had the four factors of fair use introduced, it still doesn't care about intent; it cares about how the work is used/transformed, but that's different than intent.

If you take copyrighted material and put it on a daycare mural, then you _intended_ to put copyrighted material on a daycare mural. It is not about intent to damage, it's just pure intent.

They intended to make referential work to characters in our modern common culture/mythology, as people have done since time immemorial. Some rent-seekers merely think they get to own ideas now.

Re: Illegal Prime Numbers

#85

Post by lostcolony »

snappieT wrote: post #82
lostcolony wrote: post #81

Earlier quoted context omitted.

Sometimes the law cares about intent. For copyright, it really doesn't. Hence Disney threatening to sue daycares for having Disney character murals on their walls. While we have since had the four factors of fair use introduced, it still doesn't care about intent; it cares about how the work is used/transformed, but that's different than intent.

If you take copyrighted material and put it on a daycare mural, then you _intended_ to put copyrighted material on a daycare mural. It is not about intent to damage, it's just pure intent.

What you're saying is that action = intent. Meaning intent is immaterial; the action is proof of intent. I.e., the opposite of caring about intent.

To prove intent you'd have to prove the daycare knew the works were copyrighted (i.e., the distinction between public domain Snow White, and Disney's Snow White), knew the copyright law sufficiently to know this was infringing (i.e., painting on the interior of a daycare where you're clearly not economically benefiting), and to choose to do it anyway.

Hence why copyright law doesn't take intent into account; 'ignorance of the law is not a defense'.

Re: Illegal Prime Numbers

#86

Post by SamBam »

zild3d wrote: post #76
curryst wrote: post #66

Earlier quoted context omitted.

If true, that could be an exceptionally interesting method of compression. Any volume of data can be compressed to two numbers: an offset and a length. Presuming we have a way to "stream" the numbers, rather than having to calculate the entire value.

also means the offset can be larger than the data

...and almost always is, except in the trivial cases of tiny numbers right near the start. Which is why it doesn't actually work as an encryption algorithm.

The other idea I've always liked is a length that, to enough precision, encodes something in the decimal places. Then in theory you could just cut something, say a diamond, to this impossible precision, and then measure it later. I assume that, like the Pi example, this is impractical not only in practice but even in theory, even if we had impossibly-accurate cutters and measurers.

Re: Illegal Prime Numbers

#88

Post by SamBam »

amelius wrote: post #34

If N is illegal, then could N+1 be legal or not?

I think the purpose is the intent.

If the source code to Microsoft Windows is illegal, is the source code plus one bit of padding at the end illegal? How about two bits? How about an XOR transform with some random number, which I also publish online? How can Microsoft copyright an infinite number of variations of their source code?

The answer is that they obviously can't, but all those variants are illegal if the purpose is to get back to that same source code. If you're transforming their source code, and providing some obvious or non-obvious key for how to get from your transformation to the original, you're violating copyright.

Re: Illegal Prime Numbers

#89

Post by simonh »

jasode wrote: post #17

There are several past threads of this article but this comment about how the DeCSS illegal prime was created seemed the most interesting: https://news.ycombinator.com/item?id=1045394 This might be thought of as a human version of a covert channel : https://en.wikipedia.org/wiki/Covert_channel E.g. publicizing the number and expiration date of Bill Gates' credit-card is illegal but what if we wrote a text commentary…

The number itself is not enough to have the encoded information, you also need the details of the encoding/decoding scheme. So if you had this 'commentary on Shakespear' by itself it's no more interesting than if you had an encrypted copy of Bill's credit card number but not the password to unencrypt it. Having that on a T-Shirt doesn't seem quite so outrageous though.

However if you have, or circulate this number along with an implementation of the decoding algorithm, as a reader, then you're actually circulating the credit card number itself in a much more direct way. Yes, I know they also publish the unpacking algorithms, but If I walked round with a T-shirt with the encoded file on the front, and the decoding program on the back, that's clearly a much more direct and obvious transgression.

The joke about these illegal primes is that by themselves they seem innocuous, they're just seemingly meaningless numbers, but put the two halves of the puzzle together and the problem becomes a lot clearer. Handing out an unreadable, encrypted, encoded number with the latest Marvel movie in it but no way to play it seems harmless, hand it out with a bundled video player as well and that's a problem. Now it's a viewable movie and that's a different kettle of fish.

A friend of mine got very excited when we were CS students because he found a way to theoretically 'encrypt and compress' data to extremely small sizes, until I pointed out the binary compression key would end up being vastly bigger than the 'compressed' file. Basically he was just 'sucking' the complexity out of the data file and putting it in the key. These numbers are a bit like that, a large amount of the information in the 'encoded' content is actually sequestered in the unpacking algorithm.

With the right 'decoding scheme' the number 5 is an encoded copy of Infinity War. Wow, 5 is now an illegal prime!

Re: Illegal Prime Numbers

#90

Post by gnulinux »

curryst wrote: post #66
amelius wrote: post #39

People believe (although there is no proof) that any finite sequence of bits appears somewhere in pi, making pi an illegal number :) https://math.stackexchange.com/questions/216343/does-pi-cont...

If true, that could be an exceptionally interesting method of compression. Any volume of data can be compressed to two numbers: an offset and a length. Presuming we have a way to "stream" the numbers, rather than having to calculate the entire value.

That would not be compression since the number of bits required to store offset would on average be larger than the number of bits required to store the actual data.
Post reply on HN