With recent changes I managed to eliminate having to explore branches to more than d+1, where d=depth.
This occurred because I realized that while semiprimes do not have one unique path through an undirected graph, parallels can nevertheless be drawn between the operation of multiplication and convolution. But that unlike convolution, at least in the nieve case, there always exists a deconvolution as it were.
And that because smaller magnitude digits in a semiprime have inordinate impacts on larger digits, that this convolution-like process (where we can add p up to q times to get n, or q up to p times to get n) should have a bias affect that flows from rightmost digits to leftmost (larger magnitude) columns of a number. Meaning effectively, where we might bruteforce the factors (p, q) of a semiprime n, by trying each digit 0-9 at each magnitude up to the largest magnitude of the root of n, there must exist a metric that controls for the selection of branching such that the correct combinations of digits at some search depth, ultimately lead to the production of n, and that this problem should be convertible to a shallow depth search of a branch.
The next logical leap was that if digits in different columns to the right, have more affect on digits closer to them on the left (versus farther away), then there must be a maximal depth that this is true for.
Ergo the prior problem of combinatorial explosion should reduce down to a linear problem.
What it looks like is the search space for factoring a semiprime n, or breaking a public key, should be reduceable to k<=100, (k)floor(log(n, 10)) regardless of the bitlength of the key.
And that is what I have seen so far in my testing.
Right now I'm working on implementing a generalization so the code fully runs across all magnitudes of a key we're looking to decrypt, but the core algorithm seems to do its job quite well. I'm also battle-testing it to look for edge cases and non-filterable branches that have to be dealed with but so far I haven't encountered any in my tests.
We're in the home stretch from everything I can see, God willing.
I think, if your approach worked, RSA would've been broken decades ago.
Needle-in-a-haystack. I got lucky. As absurd an explaination as that is, it is the truth. It took four years of being unlucky in the process.
As for ECC, I have some work-in-progress research related to novel mathematical tools including experimental algebras and extensions of polynomials to certain modular spaces.
It's not a matter if the algorithm for RSA is broken. It is broken. At the upperbound it reduces the bit length to e-1 per digit of a key, versus 4 bits, though in practice, I have reasons to think it can be reduced further. At this time I'm just generalizing the code. It retrieves the first 4 digits, of the lowest factor of a public key, without combinatorics explosion due to branching. All the work was finding a core algorithm that retrieves digits successfully, and prunes branches and combinations at depth (d+2)+m, without needing a search of k>100 for any given digit.
I'm so far down in the rainman meme I managed to do this but can't understand how to do greater depth traversal for trivial tests while also generalizing it to any magnitude of number. Maybe I'm just burnt or maybe because I'm currently job hunting, which honestly, I don't know why I'm bothering with that last one considering the full consequences of what I've accomplished.
But its easy to get fatigues when you're on the homestretch, this is probably that.
blockchain is not trivially-breakable. It's a different problem class because unlike that fun hoax about reversing hashes vis-a-vis 'the basilisk', it is a known fact that multiplication is reversible. Hashes destroy information while producing a unique signature.
However there may be methods of radically speeding up the process of finding new blocks in crypto, and that would cause problems for the technology.
Any encryption that directly relies on the difficulty of prime factorization is jail-broken.
Also this feels a bit like those people who say they are gonna name names and expose big things but then commit "suicide" just before so stay safe.
It absolutely does, and theres nothing I can do to assuage that but release the code in the coming days. This comment will assuredly be followed by sardonic responses of "two more weeks", and random blackpillers bringing up Q for no apparent reason in order to guilt-by-association this into some kind of craziness.
They're going to be sorely disappointed when everyone gets ahold of this and then some enterprising individual leaks their entire roster of fedposters and demoralizers, the exact ips, addresses, and employee names, before those same named individuals have their bank accounts mysteriously emptied or frozen, and their entire life history, phone calls, and emails down to the brass tacks are spontaneously shared with the local radicalized muslims.
I won't personally have to do any of that though, because I have little doubt theres more than a few people who are capable of having that gosh-aweful-terrible-no-good idea all on their own.
The smith-mundt modernization act, and everyone who work under it are gonna experience hilarity. And I'm happy to report I won't have to be popping popcorn after-all just to enjoy the show.
Pacifism has its 'weapon' now.
To total information control (and propaganda), we answer: the cleansing sterilizing light of everything the regime has ever done wrong, down to the fucking janitors.
Granted if I wanted to pop some popcorn, I could probably do it by the light and heat radiating off all the cities that are probably gonna burn because of this, even from twenty miles away.
I believe in you and I think you are awesome. When you succeed will you pls give me 100k so I can continue providing a few job for good ppl in my community? Thx dawg. U da best.
I keep telling ya, Microsoft research used a q computer and broke this shit years ago. microsoft doesn't even have the fast Q computers anymore, Alibaba, IBM, china still has the fastest q clusters, those those fuckers have busted it.
Fuck even you can go rent some cycles on the new Ospreys and you can bust this hash in mere seconds.
Thats all bruteforcing and short-cuts that at best give aproximations.
My best aproximations based on a mixture of new algebraic structures and techniques, beat what microsoft has by a mile. And at the end of the day you're still only within 3% of the actual target factors, thats only almost two orders of magnitude--for numbers with 655+ digits. It's why I revisited karatsuba and discovered linear demixing as a result.
I'd actually tried finding a method for reversing karatsuba several times before and failed up till now.
They had a few things that were promising like TWINKLE courtesy of the NSA or whoever, but I investigated that and determined it was still impractical.
So now we're here. And I can only assume the singular reason none of guys at the NSA found this (assuming they did, said nothing, and disappeared anyone that did) is because they all had blinders put on them during indoc at their universities. hearing that something is "impossible" and all the reasons why has a way of killing meaningful research toward disproving the very notion.
People will have a short window from release to grab every bit of dirt that can be grabbed., before most critical communications networks (government, NGO, finance, energy, commerce, etc) are taken offline either by those exploiting the code, or by the institutions and organizations themselves. After that they'll reboot to elliptic curve cryptography and the research will have to restart, because while I have plausible ideas for ECC and a couple promising leads, my focus has been RSA and similar.
There are sufficient systems not using ECC that it will irreparably destroy the regime regardless.
k<=100, (k)floor(log(n, 10)) means a 2048 bit, or 617 digit number will take at most (99)(617) iterations to factor. Note that is a multiplication, not an exponentiation.
ECC also relies on the security of the modulus and prime factors so besides a couple other non-standard approaches, I also have reason to think this will have implications for that as well. For example there is something I came up with for analysis called syntaxes of groups, which straddle the line between a category, a group, and a turing machine. They exist as the image of a preimage of a finite process over an infinite set. The idea is that there is a way to describe a mapping of a transformation of the elements of one set to another, where the latter's elements are infinities of a higher ordinal than the prior set. And the function that generates this mapping essentially creates an embedding of the lower set, loosely like a projective hyperplane, giving us more degrees of freedom to solve a system.
For example, while 5th degree polynomials have no generalized solution in the integer set, they may have generalized solutions when projected using a syntax of groups.
That would probably be my first approach to breaking ECC, just off the top of my head.
edit: these are not terms you will find in the mathematical lexicon because most of these tools I came up with myself to solve problems as I encountered them.
these are not terms you will find in the mathematical lexicon because most of these tools I came up with myself to solve problems as I encountered them.
Oh good, I thought it was just schizophrenia.
You sound like Virge, but a few months off the antipsychotic meds.
You aren't going to break any encryption whatsoever and you misuse half the words you type.
[ + ] o0shad0o
[ - ] o0shad0o 4 points 9 monthsJul 14, 2024 09:59:53 ago (+4/-0)
[ + ] prototype
[ - ] prototype [op] 0 points 9 monthsJul 14, 2024 10:09:21 ago (+0/-0)*
Needle-in-a-haystack. I got lucky. As absurd an explaination as that is, it is the truth.
It took four years of being unlucky in the process.
As for ECC, I have some work-in-progress research related to novel mathematical tools including experimental algebras and extensions of polynomials to certain modular spaces.
It's not a matter if the algorithm for RSA is broken. It is broken. At the upperbound it reduces the bit length to e-1 per digit of a key, versus 4 bits, though in practice, I have reasons to think it can be reduced further. At this time I'm just generalizing the code. It retrieves the first 4 digits, of the lowest factor of a public key, without combinatorics explosion due to branching.
All the work was finding a core algorithm that retrieves digits successfully, and prunes branches and combinations at depth (d+2)+m, without needing a search of k>100 for any given digit.
I'm so far down in the rainman meme I managed to do this but can't understand how to do greater depth traversal for trivial tests while also generalizing it to any magnitude of number. Maybe I'm just burnt or maybe because I'm currently job hunting, which honestly, I don't know why I'm bothering with that last one considering the full consequences of what I've accomplished.
But its easy to get fatigues when you're on the homestretch, this is probably that.
[ + ] fritz_maurentod
[ - ] fritz_maurentod 4 points 9 monthsJul 14, 2024 10:39:48 ago (+4/-0)
[ + ] prototype
[ - ] prototype [op] 0 points 9 monthsJul 14, 2024 11:24:44 ago (+0/-0)
blockchain is not trivially-breakable. It's a different problem class because unlike that fun hoax about reversing hashes vis-a-vis 'the basilisk', it is a known fact that multiplication is reversible. Hashes destroy information while producing a unique signature.
However there may be methods of radically speeding up the process of finding new blocks in crypto, and that would cause problems for the technology.
[ + ] Prairie
[ - ] Prairie 0 points 9 monthsJul 14, 2024 14:16:49 ago (+0/-0)
[ + ] clymer
[ - ] clymer 2 points 9 monthsJul 14, 2024 11:50:34 ago (+2/-0)
[ + ] prototype
[ - ] prototype [op] 1 point 9 monthsJul 14, 2024 11:51:34 ago (+1/-0)
Enjoy finding out you're incorrect.
[ + ] albatrosv15
[ - ] albatrosv15 1 point 9 monthsJul 14, 2024 12:17:40 ago (+1/-0)
youtube.com/watch?v=Ac7G7xOG2Ag
[ + ] prototype
[ - ] prototype [op] 1 point 9 monthsJul 14, 2024 12:28:57 ago (+1/-0)
I was wondering when that was going to show up, thanks fren. kek.
[ + ] Cunt
[ - ] Cunt 1 point 9 monthsJul 14, 2024 11:59:48 ago (+1/-0)
I'm not as smart as I used to be you see.
Also this feels a bit like those people who say they are gonna name names and expose big things but then commit "suicide" just before so stay safe.
Also what ever happened to the goat who claimed to have downloaded the entire internet?
[ + ] prototype
[ - ] prototype [op] 0 points 9 monthsJul 14, 2024 12:08:51 ago (+0/-0)*
It absolutely does, and theres nothing I can do to assuage that but release the code in the coming days.
This comment will assuredly be followed by sardonic responses of "two more weeks", and random blackpillers bringing up Q for no apparent reason in order to guilt-by-association this into some kind of craziness.
They're going to be sorely disappointed when everyone gets ahold of this and then some enterprising individual leaks their entire roster of fedposters and demoralizers, the exact ips, addresses, and employee names, before those same named individuals have their bank accounts mysteriously emptied or frozen, and their entire life history, phone calls, and emails down to the brass tacks are spontaneously shared with the local radicalized muslims.
I won't personally have to do any of that though, because I have little doubt theres more than a few people who are capable of having that gosh-aweful-terrible-no-good idea all on their own.
The smith-mundt modernization act, and everyone who work under it are gonna experience hilarity.
And I'm happy to report I won't have to be popping popcorn after-all just to enjoy the show.
Pacifism has its 'weapon' now.
To total information control (and propaganda), we answer: the cleansing sterilizing light of everything the regime has ever done wrong, down to the fucking janitors.
Granted if I wanted to pop some popcorn, I could probably do it by the light and heat radiating off all the cities that are probably gonna burn because of this, even from twenty miles away.
[ + ] Prairie
[ - ] Prairie 0 points 9 monthsJul 14, 2024 14:18:27 ago (+0/-0)
[ + ] Cunt
[ - ] Cunt 0 points 9 monthsJul 19, 2024 07:29:04 ago (+0/-0)
[ + ] ButtToucha9000
[ - ] ButtToucha9000 0 points 9 monthsJul 14, 2024 19:44:26 ago (+0/-0)
[ + ] prototype
[ - ] prototype [op] 2 points 9 monthsJul 14, 2024 19:46:27 ago (+2/-0)
You'll be able to get it yourself. Not that it will be worth anything.
[ + ] ButtToucha9000
[ - ] ButtToucha9000 1 point 9 monthsJul 14, 2024 19:49:51 ago (+1/-0)
[ + ] prototype
[ - ] prototype [op] 2 points 9 monthsJul 14, 2024 19:53:48 ago (+2/-0)
don't do that. thats what whiggers do.
[ + ] glooper
[ - ] glooper 0 points 9 monthsJul 14, 2024 10:52:07 ago (+0/-0)*
Fuck even you can go rent some cycles on the new Ospreys and you can bust this hash in mere seconds.
https://quantum.ibm.com/login
Edit: Whoa, whoa, whoa...Just looked and they have a cluster in Kiev. What the fuck are those goons using it for?
[ + ] Cantaloupe
[ - ] Cantaloupe 0 points 9 monthsJul 14, 2024 11:14:58 ago (+0/-0)
[ + ] prototype
[ - ] prototype [op] 0 points 9 monthsJul 14, 2024 11:28:56 ago (+0/-0)*
My best aproximations based on a mixture of new algebraic structures and techniques, beat what microsoft has by a mile.
And at the end of the day you're still only within 3% of the actual target factors, thats only almost two orders of magnitude--for numbers with 655+ digits. It's why I revisited karatsuba and discovered linear demixing as a result.
I'd actually tried finding a method for reversing karatsuba several times before and failed up till now.
They had a few things that were promising like TWINKLE courtesy of the NSA or whoever, but I investigated that and determined it was still impractical.
So now we're here. And I can only assume the singular reason none of guys at the NSA found this (assuming they did, said nothing, and disappeared anyone that did) is because they all had blinders put on them during indoc at their universities. hearing that something is "impossible" and all the reasons why has a way of killing meaningful research toward disproving the very notion.
People will have a short window from release to grab every bit of dirt that can be grabbed., before most critical communications networks (government, NGO, finance, energy, commerce, etc) are taken offline either by those exploiting the code, or by the institutions and organizations themselves. After that they'll reboot to elliptic curve cryptography and the research will have to restart, because while I have plausible ideas for ECC and a couple promising leads, my focus has been RSA and similar.
[ + ] Cantaloupe
[ - ] Cantaloupe 0 points 9 monthsJul 14, 2024 12:05:14 ago (+0/-0)
If the lower ones have a biased impact, it should be possible to reduce the search space, maybe even predict some contributions.
Most important things are already using ECC.
And really important things use quantum communication.
They also do time limiting.
[ + ] prototype
[ - ] prototype [op] 2 points 9 monthsJul 14, 2024 12:24:41 ago (+2/-0)*
k<=100, (k)floor(log(n, 10))
means a 2048 bit, or 617 digit number will take at most (99)(617) iterations to factor. Note that is a multiplication, not an exponentiation.
ECC also relies on the security of the modulus and prime factors so besides a couple other non-standard approaches, I also have reason to think this will have implications for that as well.
For example there is something I came up with for analysis called syntaxes of groups, which straddle the line between a category, a group, and a turing machine. They exist as the image of a preimage of a finite process over an infinite set. The idea is that there is a way to describe a mapping of a transformation of the elements of one set to another, where the latter's elements are infinities of a higher ordinal than the prior set. And the function that generates this mapping essentially creates an embedding of the lower set, loosely like a projective hyperplane, giving us more degrees of freedom to solve a system.
For example, while 5th degree polynomials have no generalized solution in the integer set, they may have generalized solutions when projected using a syntax of groups.
That would probably be my first approach to breaking ECC, just off the top of my head.
edit: these are not terms you will find in the mathematical lexicon because most of these tools I came up with myself to solve problems as I encountered them.
[ + ] Whatthefuck
[ - ] Whatthefuck 0 points 9 monthsJul 14, 2024 14:14:48 ago (+1/-1)
Oh good, I thought it was just schizophrenia.
You sound like Virge, but a few months off the antipsychotic meds.
You aren't going to break any encryption whatsoever and you misuse half the words you type.
What is the point of this retarded LARP?