Topic: Developing and standardizing reversible encryption algorithms and hash functions (Read 206 times)
February 04, 2018, 07:41:29 PM
Oh the baboonery.
February 03, 2018, 03:15:32 PM
Here is my response to the hostile and honest critics alike. Some people are grumpy today because they have not had their bowl of Meow-Mix today yet. Before I go on, let me say that by a reversible computer, I mean a computer that is 99% reversible as opposed to a 100% reversible computer. There will always be some irreversibility since you will need to delete information eventually at some point.
1. Symmetric encryption-decryption functions are not one-way functions and they can be made completely reversible. Everything I said therefore holds for symmetric encryption-decryption without a caveat.
2. Please look at the paper by Charles Bennett called TIME/SPACE TRADE-OFFS FOR REVERSIBLE COMPUTATION. That paper shows how one can use a process called uncomputation to compute anything you like using a reversible computer without spending too much time, space, or deletions.
3. Reversible computers will probably hit the market in the near future since they are necessary if we want to advance computing technology much further. Several researchers have even created reversible computer chips https://dspace.mit.edu/bitstream/handle/1721.1/80144/43625195-MIT.pdf?sequence=2. The reasons why reversible comptuers have not hit the market yet are
a. reversible computation takes more steps than conventional computation so the gains that one would get by reversible computation would not be enough to overcome the space and time overhead that reversible computation comes with,
b. up until now, we had no need to construct reversible computation because we could get faster and faster computers simply by shrinking the size of everything, and
c. it is kind of hard to make a reversible computer.
4. Yes. Hash functions are one-way functions. That does not mean that they cannot be efficiently computed using nearly reversible computers. Let me give an outline of how it could be done efficiently without even resorting to the uncomputation techniques outlined in Bennett's paper that I had referred to.
Suppose that A is a finite set (A is the set of all blocks being hashed). Suppose that f_x:{0,1}^256->{0,1}^256 is a reversible function for each x in A. f_x can be computed solely using reversible gates. Then the function H:A*--->{0,1}^256 defined by H(x_1...x_n)=f_(x_n)(...f_(x_1)(0)). This is the Merkle-Damgard construction. The only irreversibility that comes into play is that you may need to delete the output of the hash function later on.
5. I have read up on cryptography already.
6. The title of this room is "serious discussion." This means that you do not use words like "f***ing" and "stupid ***" and phrases such as "You're dumb" and "This post is dumb." That sort of discussion should be reserved for a room called "non-serious discussion."
7. No response. What's up with that?
1. Symmetric encryption-decryption functions are not one-way functions and they can be made completely reversible. Everything I said therefore holds for symmetric encryption-decryption without a caveat.
2. Please look at the paper by Charles Bennett called TIME/SPACE TRADE-OFFS FOR REVERSIBLE COMPUTATION. That paper shows how one can use a process called uncomputation to compute anything you like using a reversible computer without spending too much time, space, or deletions.
3. Reversible computers will probably hit the market in the near future since they are necessary if we want to advance computing technology much further. Several researchers have even created reversible computer chips https://dspace.mit.edu/bitstream/handle/1721.1/80144/43625195-MIT.pdf?sequence=2. The reasons why reversible comptuers have not hit the market yet are
a. reversible computation takes more steps than conventional computation so the gains that one would get by reversible computation would not be enough to overcome the space and time overhead that reversible computation comes with,
b. up until now, we had no need to construct reversible computation because we could get faster and faster computers simply by shrinking the size of everything, and
c. it is kind of hard to make a reversible computer.
4. Yes. Hash functions are one-way functions. That does not mean that they cannot be efficiently computed using nearly reversible computers. Let me give an outline of how it could be done efficiently without even resorting to the uncomputation techniques outlined in Bennett's paper that I had referred to.
Suppose that A is a finite set (A is the set of all blocks being hashed). Suppose that f_x:{0,1}^256->{0,1}^256 is a reversible function for each x in A. f_x can be computed solely using reversible gates. Then the function H:A*--->{0,1}^256 defined by H(x_1...x_n)=f_(x_n)(...f_(x_1)(0)). This is the Merkle-Damgard construction. The only irreversibility that comes into play is that you may need to delete the output of the hash function later on.
5. I have read up on cryptography already.
6. The title of this room is "serious discussion." This means that you do not use words like "f***ing" and "stupid ***" and phrases such as "You're dumb" and "This post is dumb." That sort of discussion should be reserved for a room called "non-serious discussion."
7. No response. What's up with that?
February 03, 2018, 02:50:28 PM
So this post has deteriorated quite quickly.
Indeed. Maybe posting something useful would be nice. "Reversible computers" would be akin to AI driven singularity. It's not here yet and not even close.
February 03, 2018, 02:46:45 PM
So this post has deteriorated quite quickly.
February 03, 2018, 01:57:28 PM
But but he's a doctor goddammit of doctorology
@op you'd be much better off reading the #trilema logs for six months, then humbly asking them your questions should they remain.
http://btcbase.org/log/
http://btcbase.org/log/
February 03, 2018, 12:20:34 PM
You're dumb.
This post is dumb.
Reversible "one-way" functions are fucking useless...
Maybe you should actually read up on cryptography before making stupid ass posts.
This post is dumb.
Reversible "one-way" functions are fucking useless...
Maybe you should actually read up on cryptography before making stupid ass posts.
February 03, 2018, 10:10:14 AM
So reversible computers are future computers which will use very little energy. These reversible computers will have to perform all computation tasks much differently than computers today perform these tasks and it usually takes more steps to compute something reversibly than to compute the same thing irreversibly (in the long term however, it will be more energy efficient to perform a greater number of steps to compute something reversibly than it would to perform irreversible steps).
So today cryptosystems such as symmetric encryption-decryption systems and cryptographic hash functions are not particularly designed with reversible computation in mind. However, in the (hopefully near) future, people will develop reversible computers and these cryptosystems need to be suitable for these reversible computers. It seems like cryptographers and cryptographic organizations should now look towards standardizing reversible encryption and hash functions so that we have these cryptosystems ready for the reversible computers when they come out. Reversible hardware could be used to increase the efficiency and security of cryptosystems. Reversible hardware will be less prone to side-channel attacks since reversible logic gates do not use as much energy as conventional logic gates. Reversible computation in general will be more energy efficient especially for algorithms that are specifically designed to be computed reversibly, and symmetric encryption and hash functions could be designed so that they take just as many steps to compute reversible (but with using less energy) as they can be computed irreversibly.
So today cryptosystems such as symmetric encryption-decryption systems and cryptographic hash functions are not particularly designed with reversible computation in mind. However, in the (hopefully near) future, people will develop reversible computers and these cryptosystems need to be suitable for these reversible computers. It seems like cryptographers and cryptographic organizations should now look towards standardizing reversible encryption and hash functions so that we have these cryptosystems ready for the reversible computers when they come out. Reversible hardware could be used to increase the efficiency and security of cryptosystems. Reversible hardware will be less prone to side-channel attacks since reversible logic gates do not use as much energy as conventional logic gates. Reversible computation in general will be more energy efficient especially for algorithms that are specifically designed to be computed reversibly, and symmetric encryption and hash functions could be designed so that they take just as many steps to compute reversible (but with using less energy) as they can be computed irreversibly.
Jump to: