4 ms·

Call me biased, but that analogy doesn't work. Explosives aren't made from physics. An understanding of the physics involved is insufficient to manufacture the

Call me biased, but that analogy doesn't work. Explosives aren't made from physics. An understanding of the physics involved is insufficient to manufacture them. They aren't just a man-made construct, they tangible objects. You can't will an atomic bomb into existence by drawing up plans for one. Cryptographic functions, on the other hand, are pure math. The concept and the implementation are one and the same. They are nothing but a mental construct built from basic mathematical principles. If even one human understands cryptography and shares that knowledge, nothing can stop the whole planet from using it. So, is there a war on math? If there is a war on cryptography (I don't think there is any more, we won that war years ago), then yes, there is a war on mathematics.

nickpsecurity 11y ago
"Explosives aren't made from physics. An understanding of the physics involved is insufficient to manufacture them. They aren't just a man-made construct, they tangible objects." "Cryptographic functions, on the other hand, are pure math." Oh, wow, you're living up to your name in this one. The bomb works through physics. The crypto works through math. The purpose and result come through a specific configuration. And, no, cryptographic functions are not "pure math:" they're instruction sequences trying to implement math on computers powered by the physics of electrical engineering. Next you'll be telling me side channels don't exist because they're not in the math. "Don't Trust the Math" https://www.schneier.com/blog/archives/2013/10/friday_squid_bl_396.html#c2056522 https://www.schneier.com/blog/archives/2013/10/friday_squid_...
- pixl97 11y ago
  >The bomb works through physics So I can take a box of physics and make a bomb? No. Bombs require chemicals and they can be regulated. >The crypto works through math. Yes, as in you don't need the computer hardware to actually do the encryption. You could do it on paper if you wanted (and had an unlimited amount of time). The encryption is agnostic of the actual hardware it runs on.
  - nickpsecurity 11y ago
    So, I can take some math equations and I have working SSL deployment? No, they require machine instructions on electrical equipment. Both can be regulated. And your model requires user to have unlimited amoumt of time? I dont know many crypto deploymemts supporting that. Your model gets even further from reality.
    - thescriptkiddie 11y ago
      You can't regulate machine instructions, dude. Are you gonna build a computer that can run any sequence of instructions except the ones that implement a cryptographic algorithm?
      - nickpsecurity 11y ago
        "You can't regulate machine instructions, dude." It's already done in a number of countries with people serving time for using it or not handing over keys. Many of those countries are implementing dragnet surveillance to get more control and detect more crypto. Or are you also myopically focused on just the math and numbers getting into the computer or pencil/paper part? Models and math don't mean crap without considering real-world requirements and implementation. In real world, those ciphertexts will go over a wire, radio, or couriered object. Two are easily detected by surveillance with randomness checks while one is harder but has human and technical means available. So, yes, you can regulate the use of a specific configuration of math on computers and/or the Internet. You can hack those systems then bypass the math using implementation bugs. You can bypass the math using its electrical properties (a la side channels). The "math" won't save you if it doesn't reflect reality and can be banned in implementation while other math stays legal. Has which you all keep ignoring. EDIT: Privacy isn't a math problem. It requires political action by citizens, math constructs, correct design of systems, mitigation of electrical issues, and tamper-resistant tech. It's a hard problem. Political part is most important as mandated backdoors or jailtime for unbreakable crypto counters all the rest easily.

nickpsecurity 11y ago

"Explosives aren't made from physics. An understanding of the physics involved is insufficient to manufacture them. They aren't just a man-made construct, they tangible objects." "Cryptographic functions, on the other hand, are pure math." Oh, wow, you're living up to your name in this one. The bomb works through physics. The crypto works through math. The purpose and result come through a specific configuration. And, no, cryptographic functions are not "pure math:" they're instruction sequences trying to implement math on computers powered by the physics of electrical engineering. Next you'll be telling me side channels don't exist because they're not in the math. "Don't Trust the Math" https://www.schneier.com/blog/archives/2013/10/friday_squid_bl_396.html#c2056522 https://www.schneier.com/blog/archives/2013/10/friday_squid_...

pixl97 11y ago
>The bomb works through physics So I can take a box of physics and make a bomb? No. Bombs require chemicals and they can be regulated. >The crypto works through math. Yes, as in you don't need the computer hardware to actually do the encryption. You could do it on paper if you wanted (and had an unlimited amount of time). The encryption is agnostic of the actual hardware it runs on.
- nickpsecurity 11y ago
  So, I can take some math equations and I have working SSL deployment? No, they require machine instructions on electrical equipment. Both can be regulated. And your model requires user to have unlimited amoumt of time? I dont know many crypto deploymemts supporting that. Your model gets even further from reality.
  - thescriptkiddie 11y ago
    You can't regulate machine instructions, dude. Are you gonna build a computer that can run any sequence of instructions except the ones that implement a cryptographic algorithm?
    - nickpsecurity 11y ago
      "You can't regulate machine instructions, dude." It's already done in a number of countries with people serving time for using it or not handing over keys. Many of those countries are implementing dragnet surveillance to get more control and detect more crypto. Or are you also myopically focused on just the math and numbers getting into the computer or pencil/paper part? Models and math don't mean crap without considering real-world requirements and implementation. In real world, those ciphertexts will go over a wire, radio, or couriered object. Two are easily detected by surveillance with randomness checks while one is harder but has human and technical means available. So, yes, you can regulate the use of a specific configuration of math on computers and/or the Internet. You can hack those systems then bypass the math using implementation bugs. You can bypass the math using its electrical properties (a la side channels). The "math" won't save you if it doesn't reflect reality and can be banned in implementation while other math stays legal. Has which you all keep ignoring. EDIT: Privacy isn't a math problem. It requires political action by citizens, math constructs, correct design of systems, mitigation of electrical issues, and tamper-resistant tech. It's a hard problem. Political part is most important as mandated backdoors or jailtime for unbreakable crypto counters all the rest easily.