are there papers that worth add to getmonero.org/resources/research-lab/? The last one was arcturus in 2020. I'd say at least OSPEAD. Anything else?

Rucknium^

ErCiccione (temporary): it would be nice if that page had a link to moneroresearch.info

> <@ghostway:matrix.org> Just a question, about what did you say that? I can't see the context, sorry. But I agree mostly. Custom kernels in cuda for example to search for new generators, can be written automagically with a config file probably

> But yea, I don't have much experience with numba but it seems cool (preferring cuda still heh)

There was a conversation above above about the tools needed to look into statistical/bootstrapped/cryptanalsysis/ML analysis.

<xmrack[m]> "tonz0fphun: As you read through..." <- Ok! I've been putting them here in the chat, but I do appreciate the notes will get lost in the noise

<Rucknium[m]> "tonz0fphun: Alternative plan..." <- Haven't coded C++ since 2017! Gave it back for Python as ML transitioned in its entirety to Python.

tonz0fphun[m]: is it possible you could summarize them here?

It can be very helpful

ghostway[m]: I've done a couple up ahead (a few days ago). I'll read a few more over the weekend, and get back to you. There's a knowledge gap I need to understand when it comes to cryptanalysis and blockchains (not my background you see)

tonz0fphun[m]: if you need updating on new features (have you used c++17?), lets start a chat?

I think most of the potential statistical attacks against Monero don't involve cryptanalysis.

ghostway[m]: I was using C++14/17, but to be honest for ML and build frameworks that interact with monero but aren't part of it, I would stay away from it

Rucknium[m]: from what I've seen so far (two papers only) they involve behavioural analysis for de-obfuscation of the ring CT, but I haven't yet seen a PoC

I will try to consider what ghostway and tonz0fphun said about numerical optimizations using Python. It might be applicable. Maybe not.

tonz0fphun: Moser et al. (2018) provides a proof of concept unless I'm misunderstanding the intended meaning. Many of their techniques no longer work well because Monero has been updated. But at least the probability distribution of the decoy selector can be improved.

Rucknium[m]: what was the title of the paper?

"An empirical analysis of traceability in the Monero blockchain." moneroresearch.info/index.php?action=resource_RESOURCEVIEW_CORE&id=15

Haven't read it yet, I'll prioritise it then

It is probably the Monero paper that is cited the most, besides the Cryptonote white paper

Rucknium[m]: That's what I was trying to find on wikindex lol

Many of the later citations are of flavor "There are major problems with private cryptocurrencies" when the techniques have been mostly rendered inert.

Hi, I have experience with cpp and other languages mostly in the field of OS's and embedded systems and I would like to contribute to the monero project. What are the current projects that I can start contribute to?

Monero currently uses the decoy selection probability distribution estimated in their Section 6.1

shalit: Welcome! That's a good question to ask in #monero-dev:monero.social

Rucknium: Ok, Thank you very much!

<xmrack[m]> "ErCiccione (temporary): it would..." <- Yeah will add that too

<ErCiccionetempor> "are there papers that worth..." <- Anything else beside this paper that worth adding?

ErCiccione (temporary): you could add the paper I wrote which was funded by MAGIC.

The research lab page currently shows only papers produced by the mrl, doesn't matter who funded them, so should be ok 🙂

ErCiccione (temporary): I don't think we should add the OSPEAD draft. It's not ready.

xmrack's paper should be added. And he should either submit it for peer review or post on eprint.iacr.org or both

isthmus said he was going to polish up a few papers, so those could go on getmonero when ready: libera.monerologs.net/monero-research-lab/20230118#c190897

sounds good

<Rucknium[m]> "xmrack's paper should be added..." <- I’ll work on this

Thanks!

<UkoeHB> "I'm thinking seraphis should..." <- I like your division and what you are proposing to audit. But I have some concerns:

1) What is considered security proofing? Getting the formal math proofs does not imply that they are safely implemented. Are we planning to have the implementation of the main functions scrutinized too?

2) Can we have the implementation of the new cryptographic functions also checked? For example, we are heavily relying on scalar divisions now and before we never had that, we have new hash functions, the new x25519 lib... I know that most of it were copied from their original libs but double checking that everything is fine may be worth?

Rucknium: can you see my PMs? matrix is broken on my end

I can see your PMs. I will try something....

dangerousfreedom: yes all new proofs and core crypto needs to be audited