01:49:19 You don't need a DAG to parallelize wallet scanning or revalidating the chain. 01:49:57 Unless your goal is to fragment consensus, where only subsets of the network are trusted to validate chunks. 01:52:53 Not sure how a multi-dimensional DAG would improve restore speed. A full node still needs to validate every block. That's what it means to be a full node. 01:53:15 (plus storing the whole chain) 01:59:22 There are a lot of design considerations. Do you have a whitepaper that lays out your full proposal? No idea how you'd make the stablecoin aspect appealing/sensical. 02:00:15 > Unless your goal is to fragment consensus, where only subsets of the network are trusted to validate chunks. 02:00:15 I’m happy to answer that. I did implement sharding and Mina style (post quantum and private) stark recursion that makes the chain not bloat with TX amount, DAG width etc. In fact, nodes can prove with only 1.8mb of data. This also required building a trustless decider which allows a proof to prove another proof. So per block [... too long, see https://mrelay.p2pool.observer/e/xta2gpELaHBHdjNr ] 02:02:53 UkoeHB: I’m a solo dev and have basically been experimenting the whole time until I found a solution. The above is what I came up with and so far it’s working on the testnet that I’ll be releasing publicly soon. 02:02:53 I do have a whitepaper V3 releasing soon with updated information on the development progress as it changes every day. 02:06:15 Without nodes knowing the entire chains data this also has privacy wins as a hacked node or post quantum attack can’t leak the entire chains data. 02:08:14 How are double spends handled? 02:10:46 UkoeHB: Every layer has nullifiers and double spend protection which makes double spend cryptographically infeasible. I ran adversarial test at every step in order to make sure that a tainted or incorrect spend is loudly rejected so there are no silent failures. 02:10:46 In fact, doing it this way offers a sound improvement, because double spends are prevented structurally rather than an add on later. 02:11:23 Yeah, but what are you using as the nullifier 02:15:03 What do you mean by 'every layer has double spend protection'? Tbh I am not knowledgeable on sharding. I only know Nakamoto and FBA. 02:17:10 Like walk me through this: what happens when two different shards receive txs spending the same enotes? 02:18:52 FCMP++ uses a fixed-order curve tree, how do the shards coordinate to construct it (without presumably losing the purported tx speed benefits)? 02:22:10 why are you wasting your time with this? :p 02:23:00 @jpk68:matrix.org: Poseidon 2 hash based deterministic nullifiers. The spend secret is derived from the nullifier. The circuit constraints force nk = H(sk) and nf= H(yada yada) so a prover can’t choose an arbitrary double spend. Since it’s deterministic one note = exactly 1 nf spend. 02:23:00 Essentially a PQ transfer tx on my chain has exactly one input and the consensus gate requires its key-8’she to equal the proof of the nullifier byte for byte. 02:24:21 @sgp_: You don’t have to believe it. You can see it when it launches. 02:25:02 I was invited here to talk to “open minded” experts who like to discuss research grade cryptography. 02:26:23 Except none of the described technology is related to Monero, a block DAG wouldn't impact wallet scanning, FCMP++ isn't post-quantum, and block DAGs don't scale. 02:26:46 sgp_: I like to understand things 02:27:04 Claiming to have copied various things from various other projects and slapped them together suggests you just had a LLM lie to you about what it's doing, while you aren't actually aware yourself. 02:27:35 If you'd like to be treated seriously, publish your actual work and not just your claims. 02:28:01 @kayabanerve:matrix.org: Well that’s the thing. I didn’t use only Monero technology. I didn’t discriminate on the source. If one project had good tech I decided to integrate it. I won’t lie like I invented the tech or wrote the papers. 02:28:23 "none of the described technology is related to Monero" 02:28:28 key word was "none" 02:29:11 "Essentially a PQ transfer tx on my chain has exactly one input" this would progressively split all funds into smaller and smaller amounts, no? Which doesn't seem practical. 02:31:19 And yes I used an LLMs help to develop it just like you guys do. But to assume that I’m an amateur is in bad faith. > <@kayabanerve:matrix.org> Claiming to have copied various things from various other projects and slapped them together suggests you just had a LLM lie to you about what it's doing, while you aren't actually aware yourself. 02:31:19 Your type of skeptical attitude is why you’ll lose talent. My code will be open source and I’m only here because I was asked to come. But don’t disrespect me like I’m incompetent and would believe an AI without building my own harness. 02:31:19 Don’t think I owe you anything, I’m just here to talk. Respect must be given to receive it back 02:31:22 Yeah I also wonder what aspect uses Monero design. 02:32:14 > just like you guys do 02:32:14 Monero is one of the communities more cautious against LLMs, as far as I've seen. 02:32:27 You should be open minded to feedback on how you present yourself. 02:32:38 You're making very big claims, not actually providing evidence, and have unclear comments. 02:32:39 LLMs are only used for debugging here and maybe suggesting solutions to annoying problems (like serialization). 02:32:56 If you want to have a discussion, you need to present yourself well enough for people to want to have a discussion with you. 02:33:11 Providing accurate details on your work, and links to it such that it can be reviewed, is necessary. 02:33:32 So, would you like to link your work or would you like to accept it can't feasibly be discussed because it only exists in front of you? 02:33:40 (and allegedly at that. big claims require evidence) 02:34:32 The actual devs of Monero use LLMs. I saw the screenshots from their discords. In fact LLMs are smart enough to hack any project. It would be a grave mistake to underestimate developing technology. Monero just recently had an exploit so no code is 100% secure. > <@kayabanerve:matrix.org> > just like you guys do 02:34:32 Of course none of my claims can be confirmed yet because the testnet or code isn’t public yet. 02:34:43 Monero has a discord? 02:34:54 UkoeHB: A private one. 02:35:00 We're not devs of Monero? 02:35:02 Wait what lol 02:35:13 I feel so left out 02:35:17 @kayabanerve:matrix.org: Hey it’s decentralized I don’t know for sure lol 02:35:19 Monero just recently had an exploit? 02:35:54 But the issue here is that we are the developers of Monero and you're telling us we aren't, so it's rather clear you don't actually know what you're discussing. 02:36:07 For reference, koe is the architect of Seraphis and I'm the architect of FCMP++. 02:36:42 @kayabanerve:matrix.org: Well technically we “all” are developers of Monero. Isn’t that what decentralization is? 02:37:12 Yes but we are actually the architects of protocol upgrades and work immediately with the people merging code which is released into what Monero is 02:37:15 Monero dev means contributing to the Monero core repo, there are also Monero ecosystem devs who work on wallets and such 02:37:30 And you're mentioning some private Discord server which sounds like script kiddies lying to you about who they are lol 02:37:38 Also, what exploit did Monero recently have? 02:38:49 @kayabanerve:matrix.org: Well I’m a big fan of your guys work! And I wouldn’t be able to do what I do without your advances in the space. All I did was try to expand upon it. 02:38:49 The goal is to not just copy the work but to expand it into something that works and is safe from quantum threats. 02:39:36 Well if you want to talk about your project, I had a number of questions unanswered above ^.^ 02:39:54 The issue is, as we've tried to say, you're making unclear and conflicting claims with no evidence, justification, or rational, but then acting as if we're disrespecting you when we push back on that. While you're welcome to have research discussions, it should be clear you need to update how you present yourself and the conversation you want to have. 02:40:40 I'd like to simply be linked to a design document/whitepaper and code repository, and also to hear what exploit Monero recently had. 02:42:27 UkoeHB: Great! I can go back up and explain but I’d appreciate a more friendly attitude from Mr or Mrs Kaya lol. 02:42:27 It seems that no matter what I say you guys won’t believe it until the code itself is released. So I see little benefit to discussing further. You can’t simply treat everyone as a scammer. There are many intelligent people out there who can benefit the community. 02:42:34 I'd also like to hear what you think this exploit was 02:44:41 UkoeHB: All code has vulnerabilities. It would be crazy to claim that your code has 0 vulnerabilities. Especially with the new growth of AI there are some things that we don’t understand on the same level. 02:44:41 https://www.reddit.com/r/Monero/s/wTopF27BA6 02:45:38 Ah p2pool. It's misleading to call that a Monero exploit, as it is only a defect in one pool's software. 02:46:59 UkoeHB: Possibly, but it still stands to say that we can’t get complacent and must stay with the times and always look to patch new vulnerabilities.