09:51:44 Secure Multi-Party Computation (MPC) is a method that allows multiple parties to jointly compute a function on their inputs without revealing their inputs to each other. This method can be implemented in Monero, to provide an additional layer of privacy for transactions by hiding the identity of the sender and the amount of the transaction. This would increase the anonymity of the Monero network. However, it would also introduce 09:51:45 additional computational overhead for nodes and may increase the size of transactions. More detailed information about the implementation, benefits, drawbacks and a comparison with the existing privacy features in Monero can be found in the .txt file. 09:51:55 * az1001[m] posted a file: (3KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/YrHKjKPdQZwvJzlmmwXDkJZv/MPC_XMR.txt > 12:31:31 gingeropolous: 104.238.221.81:18080 is also your seed node and it's down too, can you take a look? 12:31:58 66.85.74.134:18080 also seems down again, or is it still syncing? 12:35:43 it is still syncing 12:36:48 104.238.221.81 is dead 12:37:12 ok, so I'll remove 104 12:39:35 ugh, can't get a status response from 66. must be spinning the rust hard 13:00:28 No ssd..? 13:02:18 (Should have 5x4tb. Why not use one?) 13:15:03 lulz. not at that ip address 13:20:54 gingeropolous: maybe scp the blockchain from another server? 13:21:01 otherwise sync will take ages lol 13:21:10 i might switch servers 19:10:35 I want to air a possibly controversial opinion, and would be happy to get feedback 19:11:20 I think we should deprecate and remove the mine-to-use RPC payment system out of the core Monero repository and elsewhere 19:11:28 Here's some reasons: 19:11:37 1) Makes wallet development much more complicated, especially for adding new types of connection modes, since tracking credits is very finnicky. 2) Most endpoints do not require credits. 3) Bootstrap node selector does not sort public nodes by cost, so the "competing for lowest price" aspect is more or less moot. 4) Most public nodes don't require RPC payment anyways 5) The main wallet type that uses RPC with public nodes (Monero GUI 19:11:37 w/ Bootstrap Mode) will not mine and submit nonces b/c it does not use wallet2::search_for_rpc_payment. 6) An RPC payment system could be implemented as a seperate reverse proxy server on the server side and a forward proxy on the client side to keep core wallet/daemon code simpler. 7) The client payment ID signatures (especially persistent IDs) are detrimental to anonymity, especially when connecting to random public nodes. 8) It's 19:11:37 becoming increasingly more apparent that connecting to random public nodes without p2p verification can be detrimental to privacy and can introduce weaknesses where random public nodes give bad fee information etc, so the credit system makes less sense in this adversarial environment 9) AFAIK, this is not a feature that is present in other cryptocurrencies because there has not been a demonstrated need for such a system elsewhere 19:16:32 Disclaimer: this did come into my mind during development of the community node system as I noticed that around 70% of my development time was spent on maintaining that the RPC payment system remained intact for different modes of connection, which is not what I wanted to be spending my time on. The biggest problem for me personally was how the cookie + client payment ID signature acted as an extremely strong unique identifer 19:29:44 Malicious actors would simply offer their nodes for free anyway 19:44:39 Exactly: even if there was code which searched for the public nodes with the best RPC price, malicious actors could just set there's to 0 to attract those wallets and effectively DDoS wallets using public RPC 19:45:59 Also, it is entirely up to the node operator to honor those credits to begin with, so a wallet can't know the real price until it starts interacting with it. The system just doesn't make sense in a public environment 20:25:27 RPC pay has been around for a while, is there any evidence it's being used? 20:33:56 community node system? 20:34:30 ugh this node. 2023-01-27 20:32:38.302 [RPC0] ERROR net contrib/epee/include/net/abstract_tcp_server2.inl:1536 Some problems at accept: Too many open files, connections_count = 1 20:36:16 wownero recently added a donation address to get_info. i remember looking at my peer list recently, there are actually a handful of nodes with rpc pay enabled, does anyone use them? doubt it. there is this project which attempted to make use of it but seems to be dead https://repo.getmonero.org/selene/primo 20:41:00 all my limits are high. i dunno whats going on 20:41:44 gingeropolous: is this the 66. one ? 20:42:12 UkoeHB: I couple nodes have it enabled but I don't think many users use it since the majority of nodes don't require any payment for access. 20:56:09 If it's not used much, then it's hard to justify maintaining it in the core software. At the very least, refactor into separate components that aren't required to be shipped with everything that uses Monero. 21:00:22 makes no sense spending dev time on rpc pay, it's obsolete 21:02:08 The use of it for RPC access was a bad idea. I did not realize in time that the incentives were wrong for adversaries. 21:02:44 Ideally, the main use would be for external services (eg, paid access to web resources, whatever). But it's a bit complex, and nobody seems to give a shit anyway. 21:02:52 So I guess it could be removed. 21:03:40 Shame though, primo was a great potential breakthrough. 21:08:37 I think the core idea behind RPC Pay is kind of brilliant, but incentives never aligned to make it take off 21:09:12 selsta, yeah 21:09:21 If we bow to harsh reality and don't want it to continue, maybe best to leave everything untouched and just not implement it for Seraphis 21:09:54 well, its not like it has to go away forever 21:10:16 once it becomes somehow viable again, then who knows 21:10:30 gingeropolous: did you try full server restart? 21:10:43 actually i haven't 21:17:49 selsta, just remove it from the list for now. i'll make a PR when I get a better box up. 21:18:16 oh wait, its connecting.... 21:54:51 huzzah status responded: Height: 2745779/2809020 (97.7%) on mainnet, 22:06:48 moneromooo and UkoeHB That's generally my opinion: it's a great feature, but isn't really usable in the current context (public nodes) and tends to get in the way. It would work better if shipped seperately as a layer on top. Also, that way, you could come up with other payment schemes which might be easier to account for (e.g. pay a fixed amouny / byte of bandwidth) 23:45:52 I support the removal of RPC payments. Adding to the list of reasons jeffro mentioned: RPC pay requires RandomX code to be present in wallet applications, which frequently triggers false positives from anti-virus software. That has consequences beyond annoying users, see: https://libera.monerologs.net/monero-dev/20221222#c179690 (Cryptonight is also a problem, but seems to be better tolerated by AV vendors. For Seraphis, let's not use 23:45:52 mining code in any part of the wallet library.) 23:46:11 There is a large overlap between users that are not able to run their own node (due to network, storage, battery, cpu limitations) and users that are unable to mine on their machine (because of those limitations). 23:50:20 gingeropolous https://github.com/monero-project/monero/issues/8624