00:47:38 [@rucknium:monero.social](https://matrix.to/#/@rucknium:monero.social) is it possible to store less records? 00:49:51 like, only the most recent one 00:51:33 The 525 checkpoint was invalidated by a reorg before it broadcast 01:05:07 Yes. I can reduce it from 10 to 1. 01:11:15 Somehow spintered into 3 chains 01:13:55 https://testnetnode4.moneroconsensus.info/ stopped due to checkpoints (i'm mining to it now, to try to surpass the following) 01:13:57 https://testnetnode3.moneroconsensus.info/ shows 2 chains. One of them is the old/organic hashpower split off from the checkpoints. One of of them is my attacking chain, now in the lead 01:13:59 then i locally have a checkpointed node at height 534, somehow 01:16:52 The 534 node has 18 peers 02:59:58 selsta, based on the bill i got so far, its possibly to bring the cost down by a factor of 8 perhaps. But its spot instances, so its not a guaranteed thing. 03:10:40 from the top four rigs on MRR, they can somehow offer between 40 and 50 MH/s for $171 / hour 03:17:31 or $3.70 per megahash-hour 08:00:43 At those rates it would cost 17k to 51% attack Monero for 1hr 😭 08:00:48 That’s nothing 08:17:06 Yep it’s simply too easy for an attacker to get 51% for monero unfortunately. The reality is that qubic is still a problem, even though they’re basically doing everything they can to sobotage their attempts. Like doing a halving, which caused the price of their coin to collapse. But what do I know I guess. people were calling me an idiot and clowning on me for basically saying Qubic is still a threat and today proved that to be true 08:19:45 is the tl;dr that they had "could" have re-org'd +10 and they only published +9? 08:21:41 They still orphaned almost 10% of blocks today, They were definitely underreporting their actual hashrate. 08:23:03 It would not cost 17k to 51% for 1h because MRR does not have that much HR to rent 08:23:26 you would need to use cloud providers which is significantly more expensive even with spot instances 08:24:25 True, but let’s just say it’s 50,000 dollars to achieve it in reality. That’s still not very much tbh 08:25:08 it's more than 50k 08:27:30 It's more than 150k 08:27:33 per hour 08:28:05 800k/h, you can get the price down with spot pricing but not sure how much spot they have available 08:30:47 @midipoet: Yes that seems to be the case, - tl;dr: The standard 10 confirmations is clearly not enough if 40+% hashrate utilizes selfish mining. 08:31:02 and yes, you can see their current baseline hashrate on https://explorer.jetskipool.ai/xmr-tracker - it's less than 2 GH/s. Everything else is what their rent for a really short time, a few hours at most. They rent it, bring hashrate to 4+ GH/s when Monero network is naturally slower (less than 4 GH/s) due to the time of day, then they do the 08:31:02 reorgs. Last night the whole extra hashrate was not reported because it didn't go through the Qubic network (except the found blocks) - it connected directly to their Monero node. 08:31:48 They also found 4x more Tari blocks than supportxmr during that period, which is a good indication of their real hashrate. 08:32:39 Me and DataHoarder will prepare a long blog post about everything that happened in August, because some of the information can be made public now 08:33:58 Yeah sure their qubic hashrate is around 2, but if they can rent double that. Then 4gh/s is pretty close to 51% of the hashrate of monero 08:35:18 It’s just disappointing they can just rent that much hashrate, which makes their qubic pool hashrate less relevant 08:44:43 It also breaks their narrative that it's all Qubic miners doing the thing. No, it's cfb renting hashrate for PR stunts. 08:49:25 I mean yeah, but what do you expect? the dude is a serial scammer and only cares about the green. They have shown the capability to rent a huge amount of hashrate and that is a problem in itself that is not easily solved other than just hoping they run out of money or just stop, because it’s unprofitable. 09:06:01 Now I have seen numbers thrown around that to take over the monero network it would require 75 million per day by people like Charles Guillemet, so it seems up for debate how much it would cost to truely take over the network. 09:06:40 that number is to acquire the hardware 09:06:47 which, they don't 09:08:24 I was in contact with the RetuSwap administrator and he asked me to ask Woodser if it would be a good idea to increase confirmations to say 30, which would put us at 60 minutes, which is the exact same as Bitcoin. 09:08:41 My question is, would this have a bad effect on the ring signature output distribution since that is heavily weighted towards recent blocks? 09:09:44 Ah ok that makes sense. 09:11:26 An exchange, especially, does not need reorganizations causing problems. 09:11:50 currently qubic could get 30 using the same methods. 09:12:24 they drop the alt chains beforehand, but that's them deciding to do so and not because they were losing the race then 09:13:21 yesterday on discord they were seeking confirmation from exchanges setting conf count high to achieve 10+ deep reorgs (not just chains of their own found blocks) 09:16:36 That is so not comforting. The fact that they could have very well done 10 plus reorgas and only chose not to 09:18:13 alright guys, pack it up, PoW is dead. lets move to PoS 09:18:32 Asic mining ends in total capture as proven by bitcoin 09:18:44 let's make it an even easier economical capture :D 09:19:14 CPU mining ends in cheap rented hash available to attack whenever you got a bit of money to burn 09:19:30 DataHoarder, stalling PoS on monero would require ~2B dollars 09:19:41 thats a loooot more than 50k worth of hash 09:20:29 of the monero equivalent, adjusted for price depending rest of community 09:21:27 buying enough stake to attack a PoS system gets exponentially more expensive 09:22:06 i would have no problem staking XXX XMR 09:22:24 i do have a problem spending 20k on a mining rig that can be overpowered with $20 worth of rented hash 09:22:31 remember qubic has achieved enough hashrate to disrupt at times even without the direct rentals. that is coming from end users 09:22:53 just proves my point of PoW being dead 09:23:00 the specific long reorgs had further help from extra rented hashrate. but otherwise, just pay people doing PoS more 09:23:10 the ones with established stake 09:24:22 if you hold 100k in xmr stake, you are less likely to sell out to attack XMR than if you have a botnet with tons of hash that you need to turn into money as fast as possible before it gets dismantled 09:24:54 an attacker wouldnt be able to pay enough to offset the potential losses from an attack 09:25:29 with PoW the miner has no real incentive if he can just move on to a different chain 09:25:37 on different topics, here's hashrate of qubic calculated from two sample groups https://irc.gammaspectra.live/f822af98ab93c87e/marathon_hashrate_net_vs_found.png 09:26:08 black line is calculated using all their submitted shares flowing via the qubic network. these account for 640M each 09:26:15 sure, a long term botnet operator should ideally be smart enough to realize attacking monero hurts long term profits but clearly cashing out quickly is more important to them 09:26:30 numbers match what their calculations are on their own sites 09:26:32 Yeah that is true, if a person gets 51% of stake in monero. Then they’re essentially hurting themselves, because their monero would lose a ton of value. Qubic miners don’t have this problem, because they get payed in qubic. So the success of monero is meaningless to them. 09:27:06 turns out, PoS means people actually have a STAKE in the network itself 09:27:16 the dotted blue line is using the solutions that would have found a block (maybe not specifically found one) against monero difficulty 09:27:26 this is bucketed in 1h intervals 09:29:44 over that peak they found 30 blocks, in a short span (within 20-30m). Calculating over an hour gives 30-32 found blocks, where they would have found 13.6 estimated based on 12365 samples for 640M diff solutions 09:31:08 this is what sech1 refers as extra hashrate 09:31:44 this was seen again later to produce reorgs, following the pattern of extra high difficulty solutions (not matching the rate of incoming general solutions) and get ahead 10, 15 blocks 09:32:23 then these would stop/slow down, and normal hashrate would be finding a couple of extra blocks as well while monero chain got longer. once it hit their threshold, they released their altchain, however long it was then 09:33:23 X/Y axis are shit cause can't bother adding more markers, these plots will get better over time :) 09:33:59 here's the same plot over the entire duration of gathered data for these past weeks https://irc.gammaspectra.live/32166ad311982a59/qubic_hashrate.png 09:34:46 buckets of 4h, same meaning of lines. black, difficulty calculated using 640M "shares". Dotted blue, using their "high difficulty" solutions that would have found a block (not all do, for different reasons) 09:35:51 I'll try to get better X/Y markets later, https://github.com/gonum/plot makes it a bit hard to get relevant ones easy 09:45:04 Some areas on plots above might not be entirely accurate as they disabled reporting solutions on the network for a while, when they were trying to do selfish mining a few marathons ago 09:45:53 Well how would the POS system work in your view? 09:51:05 how does PoS work? 09:51:19 we have the tech with zk proofs to do anonymous staking 09:53:03 Well not all pos are the same. Like would we adopt a minimum stake like ethereum or not 09:54:01 yeah 09:54:16 put it at the sui stack 09:54:19 18.73 09:54:26 or higher if needed 09:54:43 I'd propose a stack difficulty where the stake unit increases/decreases to match a target amount of validators, personally 09:55:16 Part of that's in response to how we don't have any numbers on distribution/how many people will stake 09:55:24 (if we moved to a pos bft system) 09:55:40 i love dynamic stuff like that 09:55:59 AAVEs lending cost is dynamic like that and it seems to work very well 09:56:01 Eh, that target validators would be the constant 09:56:19 *the targeted amount of validators 09:56:38 What about a hybrid? 09:57:10 Bft? I know proof of stake 09:57:52 Byzantine-fault tolerant 09:58:06 Ah 09:58:32 barthman132: I consider including a PoS BFT finality layer as moving to a system which incorporates PoS BFT 09:59:54 Well if you can include a finality layer with a pos that would be a good idea. 10:00:05 My comments apply even if we don't move "solely to" 10:00:37 Well you’re considering doing a finality layer with the current pow anyway I believe 10:02:07 I'm advocating for a PoS BFT finality layer on top of the current PoW-advanced blockchain 10:02:43 PoS BFT finality layer on top of a PoS (as in Zano)-advanced blockchain wouldn't really make sense. 10:03:40 You'd presumably just have the finality layer also handle block production. 10:04:35 If your consensus algorithm is leaderless, for efficiency reasons, you may still want to reduce the amount of potential blocks? Instead of validators agreeing on 1 out of 500 potential blocks, due to 500 different mempools, you have then agree on 1 out of 3 blocks by requiring any candidate block have valid PoW? 10:04:57 But that's just optimizing a leaderless consensus protocol with a form of pseudo-leader election. 10:05:40 But sure, there you could use PoW, a round robin, a Verifiable Random Function, or even just apply Zano's CT PoS and call it a day. The finality layer itself wouldn't care. 10:09:51 Makes sense to me. Would anyone be allowed to stake their coins or only miners? 10:11:11 Depends on what the community decides? 10:12:26 I don't believe such a limitation reasonable 10:13:20 Well if it’s only going to be miners. Wouldn’t we want the entry level into the stake to be low, because like 98% of miners wouldn’t be able to participate if we set the minimum amount like at 32 monero or something like that 10:16:04 The point of keeping pow is decentralization but as we are seeing in literally every Pow coin just two pools end up with more the 51%. So what’s the advantage of keeping Pow? 10:16:12 Distribution? 10:17:10 ideology 10:17:11 Guess who mines the most Monero 10:17:13 Botnetd 10:17:18 Botnets 10:17:23 oh hey i love botnets 10:18:59 Isn’t the idea to make Monero unstoppable? Doesn’t seem like it 10:19:00 There’s no point at this point. If a coin like qubic can threaten us then the pow wasn’t very decentralized tbw. Hell supportxmr had 40 to 50 percent of the hashrate before qubic, so we weren’t very decentralized, even before qubic. 10:20:07 It means the blockchain continues even if the finality layer stalls. 10:21:09 What do you think would be a fair reward split in a hybrid system? 10:28:58 plowsof could we agree to be realistic for once and ban the trolls 10:29:58 Mentally ill bird 10:30:55 ACK waiting to happen 10:32:06 lol 10:32:29 it's been a while 10:32:31 were you on vacation ? 10:34:16 And who are the trolls you’re referring to? 10:34:59 monerodevareslow 10:35:10 why do you feel concerned? 10:35:15 Is there a chance we just have our dials set too low? 10:35:15 In bitcoin 6 confs (60 mins) is suggested to consider a tx final 10:35:17 Newly mined coins (coinbase txns) require 100 blocks ~16.5 hrs before they can be spent. 10:35:19 That would equate to a 30 block lock and 495 lock on coinbase coins 10:35:34 Anyone who is concerned about Monero. Ignore what mentally ill bird says. This ill person is dealing with personal issues and bird is close to ACKing. They want to censor any concerns about Qubic to prevent their feelings from taking over. 10:35:54 this so much true 10:36:08 I'm not your therapist. 10:37:20 PoW isn't dead; Monero is. Any tangible solution to the 51% threat will take months to implement in the best case scenario including DNS checkpointnig. Qubic will be attacking during this time, so by the time a solution is found, Monero will be toast. 10:38:12 This is why my username is "monerodevsareslow". All talk no action and it has been 2 weeks. Qubic is only increasing their power. 10:39:13 the longer chart for today's marathon, including known orphan qubic blocks https://irc.gammaspectra.live/313cd68a7b0b344d/marathon_monday.png 10:39:50 The qubic pool has lost hashrate, but they’re making up for it by buying hashrate for a couple hours when the monero hashrate is down a little bit. 10:39:54 People are trying to cope by labeling Qubic as a Ponzi scheme or claiming that they're renting hash power. NONE OF THIS MATTERS! Regardless of the origin of their hash power, Qubic can still 51% attack Monero. @barthman132:matrix.org is correct. 10:40:39 I'm aware that this would make the network slower but could potentially be solved at the user interface level by, for example, splitting one Monero into 10 0.1 monero chunks. This would have an effect on privacy until FCMP was released because you would be able to see that there was 1 in and 10 out, for example. 10:40:41 their pool are being ddos'd so hashrate drops 10:41:09 the rented bursts do not go to their pools, so that's not relevant 10:41:57 compared to qubic or other places, monero is decentralized both in scope and well, ownership. someone can't just force upgrade everyone's nodes same day 10:42:44 Even still they’re massively disrupting the network and they almost got 10% of blocks orphaned, so even if they can’t achieve a 51% attack they’re massively damaging monero 10:42:56 Who cares where they're getting their hash power from? The details of orphaned blocks are irrelevant. The fact that Qubic has already demonstrated its ability to reorganize 10+ blocks makes it clear. Monero is waiting for a solution, but one that won't come soon enough because the developers and community are complacent. 10:42:57 the DNS checkpointing seems to be what is the short term bandaid, and it's been tested in testnet over the past few weeks as well. As unhappy as I am as well about it not being there yet, releasing a broken system would bomb monero more than anything qubic can do 10:43:29 I don't trust ofrnxmr the scammer 10:43:35 we should have a botmoneroisovercirclejerk subreddit 10:43:46 He is taking charge of DNS checkpointing with Rucknium from Nigeria 10:43:52 Scammers are coming up with the DNS checkpointing solution. RIP MONERO 10:44:08 DNS checkpointing is in the code and has been there already 10:44:28 unrelated to any random beliefs you have about them 10:44:51 the current system is opt in, the move is to make opt out (?) and test rough edges with faster chekpointing, which was not designed for 10:44:59 I'm talking about testing DNS checkpointing 10:45:03 https://docs.getmonero.org/infrastructure/monero-pulse/ 10:45:13 "it's been tested in testnet over the past few weeks" -> this is being lead by ofrnxmr and Rucknium 10:46:48 and everyone else, plus all testnet nodes, plus all the points of presence. you could also join in, and make your own altchains there 10:47:05 How do I join in> 10:47:12 run a testnet node, it's open 10:47:36 there's no "appointed" tester, be glad you got someone to go through it this soon as otherwise in monero world it could indeed have taken months to reach consensus 10:47:57 DataHoarder, why feeding the troll? 10:48:15 he obviously do not care about your answer and your time 10:48:16 Mentally ill bird shut up 10:48:22 I'm going to run a testnode node now thank you DataHoarder 10:48:26 run monerod --testnet 10:48:27 lmao 10:48:34 https://docs.getmonero.org/interacting/monerod-reference/#pick-monero-network-blockchain 10:48:38 28080 10:48:39 i'd love to know how many qubic folk have joined this chat since this whole thing started. sometimes it feels like its +51% 10:48:45 I just blocked the troll because I know that Monero will survive this. 10:48:48 Is that the correct TCP port? 10:48:50 you want DNS checkpoint enforcing 10:48:54 also in that list 10:49:17 I think so 10:49:31 https://testnetnode4.moneroconsensus.info/ / https://testnetnode3.moneroconsensus.info/ etc. report altchains from their point of view 10:49:59 I'm not a Qubic troll mentally ill bird can even confirm 10:50:51 you probably want to run a recent version from https://github.com/monero-project/monero but unsure exactly if any specific changes have been made on monero itself yet, I think all were done around DNS infrastructure 10:51:10 add --enforce-dns-checkpointing to test them being enforced 10:51:32 12:47:57 DataHoarder, why feeding the troll? 10:51:32 at least the information on joining testnet and observing testing ongoing is probably useful for more people 10:51:59 on that topic, I should bring up my testnet node as well :) 10:52:03 Isnt the dns checkingpoint already available i recall i have it enabled in my monerod conf 10:52:26 yes 10:52:37 it's just opt out, and it had not been tested with this cadence 10:52:54 opt in* 10:53:01 by default it only notifies users 10:53:23 Oh 10:53:26 What cadence was it tested with and what cadence is it being tested with now? 10:53:47 here's the original design and purpose of this system https://docs.getmonero.org/infrastructure/monero-pulse/ 10:54:10 it's designed to fix consensus forks, not normal reorgs. so maybe one specific block if any bug arises 10:55:02 the cadence from what I saw on above posts on this channel was every 5-10 blocks regularly, which for that, DNS TTL was brought down to 5 minutes instead of hours 10:55:53 Thanks 10:56:06 I see 1800 seconds on available ones so maybe they are testing with different DNS endpoints currently, but it was discussed if you want to search on matrix history or via https://libera.monerologs.net/ 10:56:25 (here and in lab channel, but lab is not for casual discussion) 10:57:49 Yeah, I was looking for this channel earlier and stumbled upon a lab at first and realized that was definitely not the right place. 10:58:23 The lab* 11:00:39 Monero will switch to PoS if the research CCS is successful and the Monero community educates itself. Only a stupid minority support keeping PoW. Unfortunately, the time required to implement PoS is its only major disadvantage. 11:01:38 Yeah. Sadly that time is longer than shorter 11:01:54 There's still people on previous hardforks in p2pool network I see regularly 11:02:01 or people with versions from 2 years ago 11:02:22 they just never check back on it, even upgrades announced with 6 months ahead notice get people stuck behind 11:02:46 Quantum hard fork is going to be interesting then... 11:02:48 If I remember correctly, it's recommended in the P2 pool configuration sections to turn off DNS checkpointing, which seems kind of odd to me. 11:03:25 Real its not showing on my end i got it turned on 11:03:39 Reallly its not showing on my end i got it turned on 11:03:44 that has changed now, shortwave 11:03:57 in the past DNS lookups blocked the main thread so tx verification was slower 11:04:02 ./monerod 2 Dashzmq-pub tcp: 2 Slash127.0.0.1:18083 2 Dashout-peers 32 2 Dashin-peers 64 2 Dashadd-priority-node=p2pmd.xmrvsbeast.com:18080 2 Dashadd-priority-node=nodes.hashvault.pro:18080 2 Dashdisable-dns-checkpoints 2 Dashenable-dns-blocklist 11:04:03 that caused delay in block propagation 11:04:15 From https://p2pool.io/#help 11:04:19 that was fixed in monero before this, but sech1 has not updated that recommendation yet 11:04:26 I brought this up in #monero-dev 11:04:53 Is sech1 still trying to downplay Qubic? 11:04:57 on Discord 11:06:07 when? I don't think he would, he has the same data and visibility as I did on them :) 11:07:37 CFB sent me a screenshot of sech1 claiming that the Qubic hashrate is fake, stating that he had successfully attacked them somehow. He hasn't spoken much since then, so I suppose it's just more cope. 11:08:11 So does dns checkpointing have any effect on pools attempting selfish mining? 11:08:15 a note on that, we also did some trolling back :) 11:08:23 I think he posted this in the Monero Discord. 11:08:30 the sech1 backdoor is totally imaginary 11:08:45 Okay CFB took it seriously. 11:08:58 no, he takes anything to use as an excuse 11:09:10 No he didn't lol 11:09:19 that "fake" 2GH/s that showed up a few marathons ago, was real, just suppressed 11:09:47 CfB learned and no longer shows it when renting and just mines to nodes directly 11:10:03 Thanks for clarifying 11:10:04 13:08:11 So does dns checkpointing have any effect on pools attempting selfish mining? 11:10:10 it prevents deep reorgs 11:10:35 they probably can still do a couple reorgs, depending on parameters. but what is trying to get fixed is the ability to do deep reorgs to double-spend 11:11:00 note that orphaning of blocks happens even during normal conditions 11:12:13 Im intecested in understanding how discussion around workshares progressed. Did it actually go anywhere? 11:12:17 Yea but around 3-4 ? 11:12:18 also @lordx3nu:matrix.org https://x.com/xenumonero/status/1960117361240838241 I am active around these lands, but what is a member of MRL mean? I don't think I'm one :D 11:12:26 Im interested in understanding how discussion around workshares progressed. Did it actually go anywhere? 11:13:05 depends on parameters privacyx, also note blocks could come quick. the checkpoints could be set to a previous block, then that could cover it. I would suggest you check the history on this room to see what specific times they are testing 11:13:30 Orphaning typically occurs under normal conditions, but the orphan rate exceeds 10% over a 24 hour period only if a malicious mining pool chooses to engage in selfish mining. This does not happen normally 11:13:40 checkpoint flag bandaid applied on my test/main net nodes 🩹 11:13:46 xenu: i worked how to p2pool on mrr 11:13:53