Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon. Entire thread

Cryptography [PART I]

Name: Cryptography 2012-10-16 22:36

Cryptography

Name: Anonymous 2012-11-02 16:19

>>119
Fuck off and die, dipshit.  Just because you're happy to suck software patent troll dick doesn't mean everyone is.
Software patent trolls don't go around doing things for which steganography would be any kind of defense, and no amount of edgy werds changes that.

The presence of steganographic software on a computer does not prove the existence of hidden data.
That's pathetic. You think a court that would find you in contempt for not disclosing an encryption key would buy that?
For that matter, the presence of a file of apparently random data does not prove the existence of encrypted data either.

>>120
He got caught because he posted his program to a mailing list using his real identity. Steganography doesn't have shit to do with anything.

You skiddie cypherpunks are sad.

Name: Anonymous 2012-11-02 17:18

>>121
Software patent trolls don't go around doing things for which steganography would be any kind of defense, and no amount of edgy werds changes that.
For now.

That's pathetic. You think a court that would find you in contempt for not disclosing an encryption key would buy that?  For that matter, the presence of a file of apparently random data does not prove the existence of encrypted data either.
The difference is that you can convince a jury that apparently random data is very likely encrypted, but the argument that ``somewhere in the computer there might be some encrypted data but we haven't found any traces of it yet'' doesn't really work.  Unless you're in a witch trial or something.

>>120
He got caught because he posted his program to a mailing list using his real identity. Steganography doesn't have shit to do with anything.
I'm just saying that you can get in deep shit just by writing free software.

Also, supposing you have software-patent-covered or DRM-breaking software on your computer and you want to travel to another country.  Border searches are practically unlimited, and you are pretty much required to provide any passphrase to anything a border agent may find (if you want to ever see your laptop again, that is).

Name: Anonymous 2012-11-02 17:22

>>122
The difference is that you can convince a jury that apparently random data is very likely encrypted, but the argument that ``somewhere in the computer there might be some encrypted data but we haven't found any traces of it yet'' doesn't really work.  Unless you're in a witch trial or something.
Not only is that obvious bullshit (Why would you have the software if you weren't trying to hide anything? Even if you haven't actually hidden anything, no jury is going to buy that), steganography is the definition of security through obscurity. If they have your algorithm (in the form of the software you used), they have the data you tried to hide.

I'm just saying that you can get in deep shit just by writing free software.
Which is an entirely irrelevant point.

Name: Anonymous 2012-11-02 17:55

>>123
If they have your algorithm (in the form of the software you used), they have the data you tried to hide.
Aha, you don't understand steganography!  The rest of your post draws from your ignorance.

Name: Anonymous 2012-11-02 19:16

>>124
Abloobloo. If you have nothing to contribute, don't.
This thread was about cryptography. Let's get back to it.

Name: Anonymous 2012-11-02 22:00

>>117
The trick is to not require a private key, although I suppose a symmetric cipher with a plain text random key prepended to the message would do.

Name: Anonymous 2012-11-02 22:12

>>123
It is perfectly possible for steganography to use a key. But the purpose of steganography isn't to make it so they can't get the data, since that can be encrypted anyway. The goal is to hide the data's presence.

>>I'm just saying that you can get in deep shit just by writing free software.
>Which is an entirely irrelevant point.
It establishes a need for hiding information.

Name: Anonymous 2012-11-02 22:16

>>127
Why do you need someone to hold your hand through every fucking step of this conversation? Once they've found your data, which they will, you're back in the position that you yourself claimed would lead to a contempt charge for refusing to provide an encryption key.
I'm done with this argument. You're an idiot with a Hollywood/cargo cult grasp on security. Like steganography because it's cool, not because you want to pretend you need or it's good for anything.

Name: Anonymous 2012-11-02 22:18

>>126
Or just any well-documented random sequence that you xor with your data, à la OTP except not OT.
Randomly distributed data is actively bad for stego applications, though, because it stands out like a sore thumb.

Name: Anonymous 2012-11-02 23:25

>>128

We are responding to the same point. But it is perfectly possible for stenography to use a key. For instance, a scattered sequence of bits that weaves through a large collection of home photos some of which were deleted and partially overwritten with audio recordings. If the amount of encoded information is sufficiently low, how will you see the signal through the noise? Suppose they guess the exact algorithm that you used, any associated keys, and they extract the data. How will they know they have obtained the encrypted data instead of bits of random noise?

>>125
steganography is related in the it satisfies a related purpose.

Name: Anonymous 2012-11-03 5:08

>>128
Once they've found your data, which they will, you're back in the position that you yourself claimed would lead to a contempt charge for refusing to provide an encryption key.
Jesus fucking christ, how thick-skulled are you?

I'm done with this argument.
Good, get the fuck out.

Name: Anonymous 2012-11-03 5:19

>>126
The trick is to not require a private key, although I suppose a symmetric cipher with a plain text random key prepended to the message would do.
You can do public-key steganography, as long as you have some way to render the public key bits indistinguishable from random data.

Name: Anonymous 2012-11-03 6:47

>>132

I think that would be fine. Just create a fixed universal algorithm for extracting the public key from the medium. Like a facebook profile picture. But the public key must appear arbitrary, just like how it would look if the extraction algorithm was performed on an image with no key encoded in it. And throughout every step of using the public key, there couldn't be any tell tail signs of the key working correctly. If you perform the key extraction on an arbitrary image with no actual key in it, the gibberish key should function without error, as otherwise this would provide a means of identifying images without keys, which helps identify images that contain keys.

But then what exactly would the public key be used for?

Name: Anonymous 2012-11-03 10:39

>>130
You can steganographically hide data in a container that's only a few thousand (or, more realistically, dozen) times larger than the data itself, particularly if the algorithm by which you hid it is available. It's too easy to bruteforce.

>>133
Maybe you should read something like Applied Cryptography before you try to have this conversation.

Name: Anonymous 2012-11-03 12:53

>>133
That's not what I meant by ``public key bits''.  I mean the data structures inherent to the public key algorithm required to encrypt the session key.  Basically, you encrypt all your data with a random session key, then you encrypt the session key with a public key algorithm and you stick that as a header.  The point here is that while the encrypted data has normal distribution, the header might not.

Name: 135 2012-11-03 12:53

Uh, I meant uniform.

Name: Anonymous 2012-11-03 13:58

>>134
Their ability to find the data isn't the problem that's being solved here. It's the confirmation that they have found data in the first place that is being prevented. I think you are the one that needs to do some reading, kodak-san!~

https://en.wikipedia.org/wiki/Steganography

>>136
It sounds like all you have to do there is send the encrypted session key through a filter that gives it a longer and uniform distribution.

Name: Anonymous 2012-11-03 14:11

>>137
Fuck-all is being preventing. Checking whether recovered data is randomly distributed is trivial, and by your own admission they don't even need to prove they have every last bit of data; just that there's random data where non-random data is expected.
Don't insult my intelligence by linking to Wikipedia when you don't know anything about real-life cryptosystems.

Name: Anonymous 2012-11-03 14:20

>>138
Then make the encoding sparse. Only encode random data in places where random data is expected. Like the quiet white noise in a recording, etc.

Name: Anonymous 2012-11-03 14:22

>>139
Random data isn't expected anywhere except where it's deliberately put. That's the whole thing about random data.
And every stego scheme is still vulnerable to the ``problem'' you posited for simple crypto.

Name: Anonymous 2012-11-03 14:28

https://en.wikipedia.org/wiki/File:Avatar_for_terrorist.png

Le reddit face

Name: Anonymous 2012-11-03 14:30

>>140
Random data isn't expected anywhere except where it's deliberately put. That's the whole thing about random data. And every stego scheme is still vulnerable to the ``problem'' you posited for simple crypto.
So you're saying that the LSB plane of an image has zero entropy (being completely determined by the other bit planes)?

Name: Anonymous 2012-11-03 14:39

>>140
what problem did I post? I don't know who I am.

>>142
In some cases ey could be right. For instance, in a screen cap where desktop background is all #00DDDD. If you look at the image and there is a light layer of noise present in areas that should all be constant, then that is a sign. If there are objects in reality that have the same consistent color pattern all the way down to the least significant bit, then it will show there as well. There are algorithms that target regions of the image with high gradient to avoid this problem.

Name: Anonymous 2012-11-03 14:39

>>142
Sure is straw man in here.

Name: Anonymous 2012-11-03 14:42

>>144
Random data isn't expected anywhere except where it's deliberately put.

Name: Anonymous 2012-11-03 14:52

>>145
``Random'' is a word that has meaning. And that meaning isn't ``zero entropy''.

>>143
Just... stop talking.

Name: Anonymous 2012-11-03 15:12

>>146
Why don't you just stop volunteering to read what I type?

Name: Anonymous 2012-11-03 15:25

>>147
What you're typing is actively degrading an already toilet-scrubbing-level discussion.

Name: Anonymous 2012-11-03 15:27

>>147
Use a tripcode so we can automatically filter your posts. Otherwise, you're just deliberately creating noise.

Name: Anonymous 2012-11-03 15:34

>>148-149
Fuck you faggot, eat shit and die.

check my rude bump

Name: Anonymous 2012-11-03 15:40

What were we talking about again? Oh yeah.

>>148-149

<>" PROG CHALLENGE [351] [HARD] "><
Produce an algorithm that detects if a noise signal has been added to ares of the image with high gradient.

<>" PROG CHALLENGE [352] [IMPOSSIBLE] "><
Produce an algorithm that detects if a noise signal has been to a raw audio recording where the noise signal's amplitude is less than the white noise present in the recording.

Name: Anonymous 2012-11-03 15:54

>>151
Not >>148-149-kun; how are you going to measure the white noise present in an audio recording?

Name: Anonymous 2012-11-03 16:00

>>152
PROG CHALLENGE [352] [IMPOSSIBLE]
Evidently quite easily.

Name: Anonymous 2012-11-03 16:20

>>152
You know, the sound you hear when there are no other sounds. It serves as a random signal with low amplitude. Noise plus noise is more noise. Although adding two noisy signals will increase the amplitude.

>>153
Ok, then it should be easy for you to do it.

Name: 153 2012-11-03 16:21

>>154
You mistake me for FV; I am not FV.

Name: Anonymous 2012-11-03 16:24

>>155
sorry

Name: Anonymous 2012-11-03 16:27

This whole white noise thing might be more interesting if modern audio formats didn't explicitly remove noise. If you have both software that hides data in white noise and files that contain it, your imaginary prosecutors are just going to laugh at you.

Name: Anonymous 2012-11-03 16:30

>>157
if modern audio formats didn't explicitly remove noise
FLAC.

Name: Anonymous 2012-11-03 16:30

>>157
Just hide the software in the white noise. Duh.

Name: Anonymous 2012-11-03 16:33

>>157
True.

>>158
Seems too obscure. But if you are a musician that makes recordings, it makes sense to use a lossless encoding for the original.
Newer Posts
Don't change these.
Name: Email:
Entire Thread Thread List
All trademarks and copyrights on this site are owned by their respective parties. All comments and posts are the responsibility of their own posters.
- Tinychan + 2ch Mode -