-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathfreenode-##math.1019-09-06.log
122 lines (122 loc) · 9.75 KB
/
freenode-##math.1019-09-06.log
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
05:34 < aggi> in case anyone of you maths and crypto experts was interested in a snack then here it is, a cryptographically unbreakable symmetric block cipher: https://github.com/agg1/scram88
05:36 -!- Tannishpage [[email protected]] has quit [Ping timeout: 258 seconds]
05:36 < math101> I get who is at the top of number theory and combinatorics/game theory and geometry (i say nj wildberger being one on geometry)
05:36 -!- ackpacket [~ackpacket@unaffiliated/ackpacket] has quit [Quit: My puter has gone to sleep. ZZZzzz…]
05:36 < math101> But i don't know who it is for the algebra and analysis math subject areas yet
05:37 < NotTheCIA> IRC Check, do anyone Copy ?
<....>
08:16 < JizhanHuang> Hello
08:16 < JizhanHuang> Wait, still no time. -_-
08:18 < aggi> allright, another try, here it is a cryptographically unbreakable symmtric block cipher: https://github.com/agg1/scram88
08:19 < adadelta> JizhanHuang: Stuck in a singularity?
08:21 < aggi> anyone willing to discuss this?
08:23 -!- WaveHusky [[email protected]] has quit [Remote host closed the connection]
08:23 < adadelta> aggi: Not knowledgeable enough, sorry
08:24 < aggi> references to related mathematical theory is available upon request too of cause, as are known-plaintext test samples of any desired length and some simplified pseudocode
<....>
08:37 < aggi> to get back to the proposal for a moment, the relevance of the implementation mentioned is this: it is cryptographically unbreakable and 10 times more efficient and faster than any other block cipher, it's also far less lines of code and tiny IVs are utilized
08:38 < aggi> so, it might be worth the effort a skilled maths expert would have a look
08:41 < Galois> it's not worth the effort because you haven't submitted this proposal to any peer-reviewed journals or conference proceedings
08:41 < aggi> that's why i posted the github url Galois
08:41 < alphamule> AKA Too good to be true, please debunk or improve or figure out why it can't trivially be. :P
08:41 < Galois> irrelevant
08:41 < Galois> github is not peer-reviewed
08:42 -!- MoarSpaceFi [[email protected]] has quit [Ping timeout: 258 seconds]
08:42 -!- cxzxczxcxzc [~cxzxczxcx@unaffiliated/cxzxczxcxzc] has quit [Read error: Connection reset by peer]
08:42 < alphamule> Depends on the program... But not by default, no.
08:42 < alphamule> Closest you get is project leaders.
08:42 < Galois> peer review has a very specific meaning. It means that expert referees have evaluated the proposal and found it meritorious.
08:42 < alphamule> And no, you can't review yourself :P
08:43 < alphamule> So, TL;DR pay someone to publish it j/k
08:43 < Galois> it is usually not directly tied to payment, although many conferences require payment of registration to present a paper
08:43 < alphamule> Seriously though, let a bunch of cryptographers show you the ways they've defeated similar algorithms.
08:44 < alphamule> They publish
08:44 < alphamule> Shock
08:44 < aggi> i say no cryptographer will be able to break this cipher, even if he was provided with terrabytes of Zeros encrypted with it
08:44 < Galois> aggi: let me put it this way. What's our incentive?
08:44 < alphamule> "Challenge accepted" <- LOL@using psychology
08:45 < Galois> I am a super busy man. What's my incentive here?
08:45 < aggi> Galois: for the sake of God
08:45 < alphamule> There's going to be one out there that'll do it just because bored
08:45 < Galois> ok, good luck with that.
08:45 < Galois> alphamule: expert cryptographers, for the most part, are not bored.
08:45 < Galois> cryptography is in super high demand worldwide
08:45 < alphamule> Pfft, when was the last time a major new method came out j/k
08:46 < Galois> I know the answer to that. I published the last one
08:46 < Galois> It was in 2011.
08:46 < DaPinkOne> horseface, you figure it out?
08:46 < aggi> and hello to you to alphamule
08:46 < alphamule> Now network effect applies. Got to get millions of people to use it. And cool. :P
08:46 < Galois> it's getting there. Cloudflare did a series on it!
08:47 < Galois> https://blog.cloudflare.com/towards-post-quantum-cryptography-in-tls/
08:47 -!- macroprep [[email protected]] has quit [Quit: Leaving]
08:47 < alphamule> Good Ol' Decryption... wait, you meant the other thing?
08:47 -!- Lord_of_Life [~Lord@unaffiliated/lord-of-life/x-0885362] has quit [Ping timeout: 245 seconds]
08:48 -!- Lord_of_Life_ is now known as Lord_of_Life
08:48 < aggi> allright, enough said: if anyone was willing to accept the challenge then please PM and we might arrange for a scientific cooperation
08:48 < Galois> you're wasting time. nobody good will do that.
08:48 < aggi> nobody can do it
08:48 < Galois> nobody will believe you.
08:49 < Galois> anyone who is good has plenty more incentive to work on more established problems instead of rando challenges.
08:49 < aggi> it is 10 times faster and more efficient than any other cipher
08:49 < Galois> so? null encryption is even faster and more efficient.
08:50 < alphamule> I have a perfect unbreakable encryption system for web pages that's trivial to prove.
08:50 -!- sazawal [[email protected]] has quit [Quit: Leaving]
08:50 < Galois> nobody believes you, because you haven't mentioned any actual useful security claims. "My cipher is unbreakable" is not an actual claim.
08:50 < alphamule> You just need to store 10^40 bytes of random numbers on both ends. :P
08:50 < aggi> the sources are available: https://github.com/agg1/scram88
08:50 < Galois> irrelevant
08:50 < Galois> An actual claim would refer to specific techniques such as linear cryptanalysis, differential cryptanalysis, impossible differentials, and so on.
08:51 -!- WaveHusky [[email protected]] has quit [Remote host closed the connection]
08:51 < Galois> it seems you haven't even heard of these things
08:51 < alphamule> Reloading is kind of annoying though
08:51 -!- cxzxczxc1 [~cxzxczxcx@unaffiliated/cxzxczxcxzc] has quit [Ping timeout: 258 seconds]
08:51 < Galois> if you haven't even heard of these things, the chances that your cipher resists them are virtually nil
08:51 < aggi> sure, it is a matrix of simple but polymorphic LFSRs
08:51 < Galois> irrelevant
08:51 < alphamule> I was joking because it's an old finding
08:51 < Galois> "matrix" is not a cryptanalytic technique
08:51 < Galois> you haven't mentioned a single specific attack technique
08:52 < Galois> which means you don't know any
08:52 < Galois> which means you don't know how to attack cryptography
08:52 < aggi> it is 100% resistent against kown-plaintext
08:52 < Galois> which means your claims of security are worthless
08:52 < alphamule> Oh shock, if number of possible messages is the same as the number of possible keys... just use any of the reversible functions.
08:52 < aggi> only option remaining is brute force against the full keyspace supplied
08:52 < Galois> you still haven't mentioned a specific technique.
08:52 < alphamule> Well, trivially reversible. Like f(x,y)=z <-> f(x,z)=y for XOR as example
08:53 < alphamule> Oh damn netsplit?
08:53 < alphamule> It just pasted 20 lines
08:53 < aggi> i will supply any desired amount of known-plaintext encrypted, full implementation details of the algorithm
08:53 < Galois> that's not how it works. Sorry.
08:54 < Galois> ok, I'm busy enough to stop now. Good luck with your quixotic quest.
08:54 < aggi> this is exactly how it works Galois
08:54 < aggi> and i know exactly what i am doing
08:54 < Galois> tell me your list of publications and I'll judge that.
08:54 < aggi> sure, one publication was mentioned a moment ago
08:54 < Galois> please mention it again
08:54 < aggi> https://github.com/agg1/scram88
08:54 < Galois> that's not a publication
08:55 < Galois> here's my publications: https://dblp.org/pers/hd/j/Jao:David
08:55 < alphamule> Aggi: That's like Wikipedia
08:55 < alphamule> Anyone can edit their blog, too!
08:56 < aggi> enough said than Galois, no more words, how much known-plaintext would you need?
08:56 < aggi> then
08:56 < alphamule> That challenge only works if you already had it attacked and fixed the found flaws.
08:56 < Galois> I've explained why you're not doing it right. If you insist otherwise then we have nothing else to discuss.
08:56 < alphamule> No one will waste time unless you pay them.
08:57 < alphamule> Note that by 'approach', not referring to algorithm. Think about how many people submitted 'proofs' of Fermat's Last Theorem (The sum of powers one)
08:58 < Galois> Let me elaborate on that last point though. I have a quite long list of customers who pay $20000 to $200000 each for me to evaluate their ciphers. This can easily fill up all my time if I didn't want to do research anymore.
08:58 < alphamule> Hundreds... thousands
08:58 < alphamule> There's just too many
08:58 < Galois> if you're not paying you're not even in this conversation. And why would I do it for free when I can get paid?
08:58 -!- sklv1 [~sklv@gateway/tor-sasl/sklv] has quit [Ping timeout: 260 seconds]
08:59 < aggi> sure, and if you can spare 5 minutes you might have noticed some notworthy implementation details which i'll not be willing to talk about any further too
08:59 < alphamule> You can try actually using it, and then see if people will hack it. :P
08:59 < alphamule> Not sure if that wouldn't end up costing a lot more, haha.
09:00 < aggi> and one more thing: my salary is so far below any standards i had no problem with making this 100% brute force resistant almost instantly and publish that without any further notice
09:00 < alphamule> Not to discourage - it just isn't that easy.
09:00 < Galois> 5 minutes is better spent fixing your problem (that you don't know how cryptography works) than fixing your cipher (which is useless)
09:01 < Galois> but if I can't do the former then there's no reason to spend any more minutes, so I'm really out now
09:02 < alphamule> Sleeeep
09:04 < aggi> /sync