01:38:12 "Maybe someone could make the fake out selection weigh outputs from txes with higher fee more ^_^" < with same fee lvl 01:40:48 No? 01:44:51 I was told that would not be quick fix because monerod does not give wallets any information about outputs except for their age. 05:18:09 how big would a lookup table be that could be downloaded alongside the wallet 09:48:46 But then an attacker could pay slightly higher fees and increase the effectiveness of their spam attack when trying to reduce the effective ring size 09:49:58 Which sounds pretty dangerous to me 12:34:25 endor00: Whats the proposed change ? increase ringsize and change dsa to include smaller % of recent outputs ? 14:15:23 So the spammer is careful to now only put 300-500 tx in hte mempool? 14:16:02 ii am still not sure its an attack and not incognito churning/sweeping 14:24:24 Hey, it's weekend, give them a bit of slack 14:24:50 (Yeah, it's a quite noticeable change of action.) 14:26:00 if it's a deanonymization effort, keeping the mempool backlog relatively low makes sense because you still want people to use the network and this leaves room for other min fee transactions to go through pretty quickly. otherwise it's just polite 14:28:26 yeah but if this is the goal they are stupid to have started by pumping so much stuff into the mempool 14:28:58 gradual increase would have had nobody screaming about an attack 14:30:00 nobody said they were competent :D but yeah -- although a sustained 10x transaction volume increase with no corresponding price movement or exchange volume would be.... interesting 14:30:14 even if over a longer period 14:32:50 the main thing this attack has achieved is make me annoyed at people calling for 20 cent transaction fees 14:34:31 Right. 14:35:11 I will say xmr is in a rather unique position that it has financial incentives to spam the network i.e. get paid for providing an XMR tracking tool 14:36:01 of course fcmp would stop this dead 14:39:54 can we not talk about fcmp because its not coming for few years atleast, work with what we have right now 14:40:20 :x 14:41:46 >a few years atleast 14:41:55 is seraphis really doing this badly 14:45:21 from past experience, until it hits testnet there shoud be no eta hence years 15:26:12 When I have mentioned FCMPs on twitter, some have asked how they can donate to the effort. Not that throwing money at it would speed up the process (maybe support those working on it though). Just something to consider. 15:48:56 welp, back up to >3k 15:49:00 (in mempool) 16:22:35 Yay 16:22:48 Has the wallet update gotten released yet? 16:24:00 CLI is tagged and can be built, GUI is not tagged yet. no binary releases yet, hopefully later today? other wallets should be able to start rolling out updates with CLI tagged though 16:28:22 Ok 16:28:40 I wonder why it stopped for that time though 16:29:06 Maybe someone turned it off while they slept and then turned it back on when they woke up? 16:38:50 Well, making it the pool stays at a more or less constant 500 over hours and neither fills up further nor depletes IMHO needs some automation software, that would be hard and boring to do "by hand". So I guess they can put the dial on their system more or less to whatever they want regarding the size they want to keep the pool at automatically (of course with some upper limit) 19:04:44 I think the spammer is getting smarter. If their target is to make median block size grow, they don't need _all_ blocks to be full. Up to 49% blocks can be empty, all they need is to push that median higher and higher. Right now 15% of blocks are below 300k 19:05:30 So they're probably trying to save on the fees now by leaving some blocks empty (without their spam) 19:06:15 oh fun 19:06:39 median basically ignores 49% of the smallest blocks, so they can be empty 19:07:35 but getting to 49% is dangerous because it can slip over 50% and then the median will be reset to 300k - back to square 1 19:08:03 huh, yeah, realizing the median is not necessarily a smooth function at all 19:08:42 that's quirky 19:08:55 Which means they're probably running out of money 19:09:02 What would be the advantage of increasing the penalty median while leaving as many blocks empty as possible? 19:09:18 less fees with the same end result 19:09:23 if they're running out of money then it was never an adversary worth worrying about anyway heh 19:09:34 broke asses 19:10:01 even if not running out of money, no reason not to be efficient if u plan to go long term with it 19:10:31 Not sure I follow. What is the end result you are referring to? What is the consequence they are targeting? 19:10:33 either running out of money, or minimizing expenses for the long run 19:10:48 They're targeting higher block sizes, for whatever reason 19:13:09 and of course they're controlling more than 80% of outputs now - not enough to get all decoys in transactions though 19:13:53 I would think higher allowed block sizes are not practically relevant if they are not filled. 19:13:53 If you intend to fill larger blocks, why would you not fill the smaller ones? 19:13:57 maybe they aren't targeting block size, maybe that's jsut a side effect of wanting 80+% of transactions 19:14:07 but they you would want to fill all the blocks 19:14:14 so ig it is block size? 19:15:17 or maybe their target is to have a sustained transaction count/day, but blocks just keep growing because of how median works 19:43:09 they already let the median reset at some point 19:49:36 that was in the first day of the spam, but now it's continuous and doesn't reset 20:18:10 The larger the blocksize gets, assuming the traffic that is not from the entity remains as it has been, the greater the percentage of txs of each block the entity can control 20:53:23 So maybe back to 2016 or so, with a ringsize of 4. Maybe sgp has still some calcuations around that he used back then to motivate to switch to a larger ringsize ... 22:14:30 if I were the attacker, I would have started gradually, slowly inflating the transaction volume a long ago, then do a high-visibility performance like we had now, and once it's over, have people believe it's no longer happening. the more content they are, and the less they prioritize switching to full-chain membership proofs, the better for me 22:16:53 How do you know, if they havent already done it and dont want to keep scaling it up ? 22:35:26 not sure I understand your question