/crypto/graphy General

It took me 2 years to find those resources... Books Jeffrey Hoffstein, Jill Pipher, Joseph H. Silverman - An Introduction to Mathematical Cryptography Koblitz, Neal - A Course in Number Theory and Cryptography Dan Boneh and Victor Shoup - A Graduate Course in Applied Cryptography - https://toc.cryptobook.us/ Jean-Philippe Aumasson - Serious Cryptography: A Practical Introduction to Modern Encryption Alfred J. Menezes, Paul C. van Oorschot, Scott A. Vanstone - the Handbook of Applied Cryptography - http://cacr.uwaterloo.ca/hac/ B. Schneier - Applied Cryptography Links to Libraries and Zines https://github.com/isislovecruft/library-- https://github.com/fdiskyou/Zines Other Resources: https://cr.yp.to/ and sub domains https://pqcrypto.org/ https://www.schneier.com/ https://crypto.stackexchange.com https://eprint.iacr.org/ https://www.iacr.org/archive/ https://arxiv.org/list/cs.CR/recent https://www.ieee-security.org/ https://cryptovillage.org/ https://www.reddit.com/r/crypto/wiki/index Youtube Links: https://www.youtube.com/playlist?list=PL1xkDS1G9As7E_fPaLaFchq1a27I9a5tO https://www.youtube.com/playlist?list=PLAwxTw4SYaPnCeih6BPvJ5GdqqThGcWlX https://www.youtube.com/watch?v=GSIDS_lvRv4&list=PLlhlTo93Q0hDr0OrJ9qVj5RI1qxbCQK2S https://www.youtube.com/channel/UC1usFRN4LCMcfIV7UjHNuQg/ https://www.youtube.com/channel/UCYO_jab_esuFRV4b17AJtAw ...be kind, share knowledge.

I learned some basic ciphers from auditing a college class but I missed a lot of days. It's amazing how crude yet effective most cryptography is, when you get down to the mathematical fundamentals. The technicalities and practicalities, as well as dealing with quantum computers, is the hard part.

advent of books?

>>16090 >>16087 I'm still reading the Koblitz's book, but starting a conversation on this topic might help me understand cryptography at its best.

What is a good programming language to use for cryptographic purpose?

>>16259 C++ is what I see most often when looking through projects in academia. Although, haskell is also a popular selection for mathematicians, just in general.

>>16090 It surprised me how much more complicated symmetric schemes like AES tend to be than RSA or what have you. Asymmetric is essentially just cypher=pow(clear, pubkey) in whatever field you're operating on. AES on the other hand has multiple iterations and s-blocks and all kinds of weird padding and partitioning logic. I learned this all from a crypto class run by my undergrad's math department so they didn't go into implementation at all but later I realized a lot of the seemingly convoluted decisions they went into the AES standard come from the necessity to implement it on asic chips back in the early 90s or whenever it was invented.

"Roll your own crypto attacks." ~ https://twitter.com/SchmiegSophie/status/1210982785994780672 "[...] generally, we need more publication/sharing of failed attack attempts" ~ https://twitter.com/veorq/status/1211919319862579200 "Cryptography is the attempt of solving social problems with math. As such, a solid understanding of sociology and philosophy is as important as an understanding of mathematics when attempting cryptography." ~ https://twitter.com/SchmiegSophie/status/1212790611826233344

https://zkp.science

>>16262 symmetric crypto is designed to be *much* faster. doing RSA on 5368709120-bit numbers would probably never terminate in your lifetime.

>>16087 https://toc.cryptobook.us/ has been updated

Here is a few

>>16259 Plankalkul

https://sites.math.washington.edu/~koblitz/crlogia.html http://laser-calcium.glitch.me/

>>16614 https://www.robertxiao.ca/hacking/sarah2/

>>16087 >neal stephenson sage because OT.. i haven't read cryptonomicon but anathem is probably my favorite sf book. if i had to make up a subgenre for it i'd call it "speculative epistemology" or something. diamond age is good too. really cute and heartwarming, which is something you don't get too much in sf.

Mark Stamp, Richard M. Low - Applied Cryptanalysis_ Breaking Ciphers in the Real World-Wiley-IEEE Press (2007) https://cryptohack.org/challenges/general/

>>16087 >>16630 I'm reading Cryptonomicon right now! I got the book from my library before quarantine began... maybe I own it now? It's surprising how undated it feels, especially for a book centered around technology published in the '90s. I'll try Anathem next.

Hi I looked into cryptography and I'm doing some exercise, but how do I keep myself up-to-date with topics in cryptography? Is there anyone to talk with about cryptography? I feel lost after reading those books.

>>18608 They looks more like Security imprinting than Cryptography. I mean they are Okayish. Some I never heard of and dunno, but usually Schneier and Hacker's Playbook are a good _starting point_. I'd really like if someone knows how to continue from there.

>>18596 if you're getting lost because of the math stuff look into finite field theory. any undergrad abstract algebra textbook will get you started. just read the sections on groups, cyclic groups (especially the fundamental theorem of finitely generated abelian groups), whatever combinatorics they include, rings/fields (especially the parts about zero divisors and integral domains), and then finite fields (i.e. cyclic groups of prime order). there's definitely more math involved than just that, but at least you'll be able to understand RSA. if you want my textbook recommendation: Fraleigh's "A First Course in Abstract Algebra".

>>18637 I know how RSA works, probably not all RSA's weakness. I've read already some good books on algebra like Koblitz's one. A book that's helping me a lot is the one about Cryptanalysis. I think a book I was looking for was those, that talk about ellipic curves, lattices, zkp and so on. Nigel P. Smart - Cryptography Made Simple ISBN 978-3-319-21935-6 ISBN 978-3-319-21936-3 (eBook) DOI 10.1007/978-3-319-21936-3 Mark Stamp, Richard M. Low - Applied Cryptanalysis: Breaking Ciphers in the Real World ISBN 978-0-470-11486-5 Jeffrey Hoffstein Jill Pipher Joseph H. Silverman - An Introduction to Mathematical Cryptography ISBN: 978-0-387-77993-5 DOI: 10.1007/978-0-387-77994-2 e-ISBN: 978-0-387-77994-2 === but anyway, after reading those, where should I go? I dunno how should I read papers I have to, as there are books of papers like CRYPTO (Advances in cryptography) that has a lot of papers, but dunno...

>>18719 Yeah I would suggest just reading papers like you said. Just pick an encryption technology you rely on, like double ratchet or whatever, and dig through the work that produced it.

https://cryptohack.org/ I'm doing a run and trying to do it in Haskell. Both for learning more livraries in haskell and crypto.

https://cryptopals.com/ Also if you get a RSA factoring challenge I've found that gap in sagemath (or just gap itself) seems to be good at finding the best factoring algorithm ng = gap(n) p,q = ng.Factors() p,q= int(p), int(q) https://doc.sagemath.org/html/en/reference/interfaces/sage/interfaces/gap.html Also here's all the algorithms it offers https://www.gap-system.org/Manuals/pkg/FactInt-1.6.3/doc/chap0.html

>>18722 ELI5: How should I read a paper? They seems very complicate and I don't know what should I look for for compensate or expand my knowlege on topic.

>>19251 most papers start with an introduction section, and they will cite other things which the paper is about. Pick a paper about something you're interested in. Then look for through the references in the introduction to find something that looks like a "survey paper", they will usually say someting about being an 'overview' or a 'survey' in the abstract. Read the survey paper, and follow the references about stuff you don't already know.

A bunch of Germans book recommendations

>>19369 ???

>>18596 https://www.metzdowd.com/ and https://arxiv.org/search/cs?query=cryptography&searchtype=all&abstracts=show&order=-announced_date_first&size=50

>>19252 How do I continue then? There are a lot of references and the graph keeps growing. What should be the criteria to pick and read a paper other than my curiosity?

>>19771 Curiosity is the only worthy criteria. If it looks overwhelming, cover part of it up and start with what you can see.

http://cr.yp.to/talks.html

>>18637 I was trying to approach this path, also checking with cryptohack.org but there are thing I can't find in books and can't google for the right paper. Ie. RSA where the prime number is generated by adding a chunk and checking if it's prime, otherwise shifting and adding... and finally you have a prime composed by those OR and Shifts... But how do you recover the prime?

Does anyone know of an audited implementation of plain threshold-1/unlinkable/untraceable ring signatures? While the original RST implementation looks straightforward, I'm concerned about the possibility of introducing subtle errors by accident - especially since there's not much to compare it with/validate against.

What's the mathematical background for cryptography? What book would you suggest to get that mathematical background?

>>20916 Well good to have basic university level Mathematical Analysis that is usay being taught on Computer Science, plus Information Theory and Statistics, some Algebra (especially if you want to go into quantum cryptography). I think that's a good start.

How does cryptography work on non-latin languages? I mean how would someone go and encrypt a Kanji or Arabic writing?

>>21514 I mean, it's operations on the numbers that translate to the character on the screen. So it works the same.

How can I get a crypto job if I am a third world soykafter (Philippines)? Should I apply to grad schools in Europe?

>>21529 From what I seen it depends. I know someone that was a college drop out and is doing well in the crypto field. It all depends.

https://soatok.blog/2020/11/14/going-bark-a-furrys-guide-to-end-to-end-encryption/

>>22254 Why is he using Ed25519 keys with X3DH? I thought the purpose of X3DH was to not use signatures? It looks like he's using birationally equivalent X25519 keypairs,but still passing Ed25519 keys around.

Ohai! I saw this site in my referrer log :3 >>22254 Thanks for sharing this here <3 >>22275 That's a good question; one that I thought was worth adding to the Q&A section at the end. https://soatok.blog/2020/11/14/going-bark-a-furrys-guide-to-end-to-end-encryption/#why-ed25519-keys-x3dh

>>22279 Welcome! Thank you for writing a nice guide. You don't actually have to list your name/email to post. In fact, it's rather unusual to do so. Most just post anonymously. The names you see above posts are just randomly pulled from a list when you leave the name field blank.

>>22281 Ah, yeah, I haven't been on a chan board in years. Y'all seem friendlier than the ones I'm familiar with. Thanks for the kind words.

Are there any hands on crypto/programming books? By this I mean a book that doesn't just tell you to import a library and use it but actual goes through implementation and code.

>>22407 No books in particular come to mind today, but this might help: https://soatok.blog/2020/06/10/how-to-learn-cryptography-as-a-programmer/ Also, David from Cryptologie.net is writing a book on cryptography that might be worth checking out: https://cryptologie.net/article/504/why-im-writing-a-book-on-cryptography/

>>22418 >I will not roll my own crypto unless I'm ready Thanks I think for know I am going to go with http://toc.cryptobook.us/ Seems like a good intermediate step to take.

>>22279 >>22283 >>22419 Begone f*rry, and stop spamming your pornography.

>>22423 Dude, chill. Don't be a jerk. Furries have a weird thing going, sure, but they're still people.

>>22425 I'm just asking that he doesn't post his obsenities on this site, not denying his humanitu (in fact, quite the opposite).

>>22423 That furry knows a thing or two about cryptography. I'd rather have furposting with actual competency than the /g/ exiles memeing soykaf they don't understand that we usually get.

>>22425 >but they're still people. debatable

>>22425 >but they're still people Just as I find annoying people that are unable to post a message without including an anime character's picture, so does I hate what that furry is doing now. >>22427 >That furry knows a thing or two about cryptography Sure. But that doesn't give him free pass to also post furry images in every post he does.

>>22423 I would never. :3 >>22426 That's the thing, though: It wasn't me. I just posted the heart emote because it reflected my emotions at the moment. If folks are asking about technical stuff, and I have a blog post that's relevant to their question, I am going to point them towards it. Being helpful > being accommodating to people who don't like furries. >>22428 UwU >>22431 You're barking up the wrong tree.

>>22419 Cryptography I is pretty good! Everyone's waiting for Crypto II to come out (some day?).

>>22437 >It wasn't me I vouch that, it was me (not furry) I just like his blog post and wanted to show appreciation. Don't hate on him (I assume) he's really smart.

>>22431 Why are you on an imageboard if you hate images? Thanks OP & >>22419 for recommendations.

>>22447 >Why are you on an imageboard if you hate images? I don't hate images. I hate people spamming it as if it it was their avatar.

>>22437 >>22441 > That's the thing, though: It wasn't me. I just posted the heart emote because it reflected my emotions at the moment. I'm reacting to >>22419 and >>22283, and don't care about the article. >>22449 This.

I've published a few papers in top crypto conferences. Rosulek's book is a good place to start for the more theory side of basic crypto. Not the mathematics as much mind you.

>>22407 This might be what you are looking for

>>22554 this book looks good. Thank you! (not him, just bypasser)

>>22554 >honeypot.pdf >/Subtype/Form/Type/XObject>>stream Why does the PDF you posted have hidden forms in the binary?

>>22567 fuarrrk, is this real? I'm too much of a noob to know and I had already downloaded it.

>>22567 Always view PDFs with gv, never use a PDF viewer with any sort of internet connectivity, scripting or drm.

Is using blum blum shub still a reliable way to generate secure random numbers?

>>22769 Yes. It's one of those techniques that isn't widely used because it is inefficient on modern hardware, not because it is insecure.

>>22567 >hidden forms in the binary How do you check for this? Ideally, on many files at once.

>>22769 The BBS psudorandom generator relies on the hardness of factoring integers, so if possible I'd work with something else in the interest of future security.

>>22821 Sorry, I should have presented an alternative. IIRC you can chain block ciphers in counter mode, but be careful to swap out the key every once and a while.

>>22567 What does this mean? I thought it might be too on-the-nose to do this for a security pdf. That being said, we don't know if >>22554 just reposted this. >>22776 Grep does the job: fgrep '/Subtype/Form/Type/XObject>>stream' (filename) . Add a -a if you want to see the line.

>>22823 >Grep does the job Notepad++ has a "find in files" function, is that good enough? It's the only string-finder-whatever which I've gotten to work.

>>22823 >>22830 Well, Notepad++ seems to work. Any other suspicious "hidden forms" to search for?

>>22609 >never use a PDF viewer with any sort of internet connectivity, scripting or drm. Are there any PDF viewers without connectivity/scripting/DRM on Winblows?

Never use winblows.

Did you say n00b friendly lain?

This one is a nice writeup on LLL and "custom" Coppersmith for what I've understood. https://priv.pub/posts/pbctf-2020/

After finishing up the textbook: Introduction to Mathematical Cryptography, what can I do next? Would it also help me on cryptocurrencies?

What are the prerequisites for something like Serious Cryptography? Will high school math do? If not, what do I need to learn prior to learning crypto?

What is the best way to learn the basics of using cryptography? I figured I'd rather get used to GPG etc first before I start getting into the math behind it with one of the books from the OP e.g. Serious Cryptograhy

>>22776 https://www.decalage.info/exefilter ? It doesn't remove jbig2 by default, that can be changed in config.

>>24058 Yes, it should be fine. You can always brush up on math as you go through the book.

>>22880 I love these, I remember the one on regression analysis. I should really try learning from one of them some time.

>>24057 Cryptocurrencies are also a lot of byzantine analysis. I don't know of any good books on it, but consensus protocols are something you may want to look into. (of course also zero-knowledge proofs if you haven't already looked into those). Also, I'm thinking about making a cryptography blog and I'm curious if you all had any topics you'd like to see.I already have a few topics in mind, but any ideas of what you'd look for in such a blog would be great! I'm not really an expert on things like IO, homomorphic encryption, or lattices, so I probably won't write on those for quite a while.

>>24170 I want interesting pen and paper ciphers. Sarah2 is really helping me understand a lot of other stuff like aes. If you could create some interesting and reasonably decent stuff like that it'd be killer. Sarah2 and Solitaire are the two that come to mind. If you made a little JavaScript applet to go with each post you'd have something really amazing.

https://soatok.blog/2021/02/11/on-the-toxicity-of-zed-a-shaw/ The teardown of Secure Remote Password v6a was pretty neat in this post

>>22880 Just what I've been needing, thanks.

https://cryptohack.gitbook.io/cryptobook/

I've been thinking about how Freenet might be surpassed since it's been over twenty years and nothing has really addressed the flaws in its design. Chiefly, how convergent encryption doesn't provide any real deniability for files you happen to have all the pieces for. This is a problem for long term efforts to preserve controversial content (not a euphemism for CP given the recent surge in government censorship, public puritanism, and the death/decay of golden-era pirate sites) because it puts all the risk on the people who actually make content available to everyone else. So far I've come up with two sorts of general ideas on how you can have addressable content without that very addressability putting hosts at risk. The first is if you had some sort of secret sharing protocol such that when provided with a key k and the product of any number of unrelated shares (as well as the t related shares) you can reproduce the corresponding secret, then you could distribute the shares to hosts under the name of the first couple of bytes of the file hash. That way users seeking to download would simply download all of the shares in the bucket that their target hash falls under to reconstruct the secret without revealing the other secrets that might be hidden in the pile. This gives: >Protection from known ciphertext attacks, since the shares are nondeterministic >Moderate protection from offline brute-force attacks, since the adversary would have to download all these shares and perform the potentially expensive re-assembly computation for as many times as it takes to get a "hit" >The above is exacerbated if the shares were indeed split between multiple hosts, since it is unclear exactly how many of the shares a given host has if it takes two or more to reconstruct. The other idea is to instead use a distributed key and have some sort of transfer protocol that hides which server has the file while simultaneously assembling the key and decrypting with it, This way, performing an offline attack requires seizing a number of servers instead of the singular assuming that the only way to know which file corresponds to the key is to test each and every file, since otherwise the servers would know of their "guilt". Both these schemes are stupid expensive in their own ways and therefore probably impractical even if they were possible, but I really think there's value in brainstorming this stuff since we're long overdue for something better. Maybe someone more knowledgeable could pick a good idea up or pass it along because I've yet to find any new research in this area. Also, my literature review brought up Message-Locked Encryption, which seems promising on this front since it appears they've managed to work non-determinism back in.

>>27456 I think that even just obfuscating what different links are sharing to one another like suggested [here][0] under NOISE protocol might be helpful. Onion/Garlic routing would help a lot too. [0]: https://datprotocol.github.io/how-dat-works/#future-of-dat

AES cryp-tana-lysis https://www.davidwong.fr/blockbreakers/index.html