00:41:19 hi 02:33:02 To be honest, I have always dreamed of something similar to this, but he can't run all the node projects on a Raspberry Pi. If there are any vulnerabilities discovered in the ARM architecture, the entire network could be compromised. 02:33:25 ARM architecture, specifically? 02:33:51 not the CPU that raspberry Pi uses? 07:32:11 and with XMPP stanzas, too? > Cindy: activitypub but using matrix protocol, via JSON-XML 10:43:02 Maybe I was too tired to understand what it says. 10:44:11 Raspberry use ARM based CPU > not the CPU that raspberry Pi uses? 10:47:52 "vulnerabilities in the ARM architecture" what the heck are you talking about lol 10:58:44 wait 10:59:03 the proposal looks like it was written by AI 10:59:10 or at least in a style of AI 10:59:20 "This is not a research toy. It is a pragmatic path to a decentralised, censorship resistant market[...]" 10:59:27 oh MY GOD 10:59:32 IT'S NOT X, IT'S Y 11:01:03 i mean it's a very shallow and empty proposal 11:01:07 lots of words, little content 11:02:16 complimentary mdash in the title 11:02:26 hey, i use those 11:02:40 been using them before LLMs did! 11:02:44 lol 11:03:17 also the file has headings... without the # or at least the colon after 11:03:34 like "Who" or "What", "The Proposal" 11:03:57 which makes me think that this was specifically copied from a webpage rendering and pasted over 11:04:04 probably from chatgpt 11:04:06 @helene:unredacted.org: Emdash users are truly the people hurt the most by the rise of LLMs. 11:04:16 The ARM architecture is supplied by ARM Holdings, a private company that designs it. They do not release the license as open source. My point is that, as I understand it, the network would be powered by Raspberry Pi, and I think it is not a good idea to put all your eggs in one basket. 11:05:52 Powered by chatgpt* 11:06:36 Hello there 11:08:47 @ofrnxmr:xmr.mx: no correct by an assitant for helping me because I have a aphasia 11:09:48 Assistant using chatgpt****, sorry 11:10:39 Medical assitant using ChatGPT would be better 11:10:50 https://0x0.st/KTPn.png 11:11:22 @fartbubbler:matrix.org: i'm one of the most oppressed minorities, right after gamers 11:11:44 wtf is 87% mixed 11:11:52 @helene:unredacted.org: You're israeli? 11:12:11 xd 11:12:15 Cindy: Most text is edited 11:12:29 No she from Gaza 11:12:44 well, same about x86, which is exclusive to both AMD and Intel...? > <@octagonsupreme:matrix.org> The ARM architecture is supplied by ARM Holdings, a private company that designs it. They do not release the license as open source. My point is that, as I understand it, the network would be powered by Raspberry Pi, and I think it is not a good idea to put all your eggs in one basket. 11:12:44 they provide licenses to the instruction set architecture (ISA), and separate licenses for those who want it of silicon designs. a lot of companies do their own designs, like Amazon, Google, Samsung, Apple, Qualcomm, etc. 11:12:45 then she is not oppressed. 11:15:02 @ofrnxmr:xmr.mx: More how the fuck she have access to Internet 11:15:12 if this proposal was AI-generated, i'm gonna be sad 11:15:22 because i thought they spent 18K on a server ;_; 11:15:24 but I don't see how a buggy CPU would be problematic to the network, there's plenty of those already (and I'd be more concerned about driver bugs, kernel bugs, ...) 11:16:23 @helene:unredacted.org: My point was to not put all your eggs in the same basket. 11:16:43 a buggy CPU can be mitigated with software 11:16:58 with some performance cut 11:17:06 well, currently they kinda are in the same basket by your standards, almost all Monero nodes run on x86_64 CPUs 11:17:20 There is risk-V bkard that start to benn cheap 11:17:24 depending on how severe it is 11:17:43 like if it's a bug on the speculative branch predictors, mitigations are fine 11:17:56 it's a bug on the ALU................ yeaaaaaah 11:18:03 if it's a bug* 11:18:11 What is this room? 11:18:12 like the pentium 4? :P 11:18:24 Talking about menero? 11:18:29 the pentium float inaccuracies, yes 11:18:38 hanzemito: talking about monero 11:19:15 tbh, even the buggy pentium float can be mitigated with software floating-point 11:19:27 using the integer ALU 11:19:38 if that's buggy too.. you might aswell throw away the CPU 11:19:51 Hmm... so you're talking about monero working system right? 11:19:54 It think the pentium 4 didnt have microcode update of the last spectre vulnerability 11:20:04 I'm sorry, my english is too bad. 11:21:21 Me either and + aphasia problem, its why I used LLM. 11:21:56 i try my best to write long essays by myself.. even taking more than a week to do so 11:22:07 just for people to use LLMs anyway :P 11:23:05 So is this room having any rules? 11:23:15 Like no offensive. 11:23:21 Yes > <@hanzemito:xmr.se> Talking about menero? 11:23:26 @hanzemito:xmr.se: Yes 11:23:27 yeah dont talk about qubic 11:23:32 @hanzemito:xmr.se: Yes 11:23:32 yes no offensive 11:23:37 @octagonsupreme:matrix.org: Yes 11:23:43 don't say racial slurs 11:23:54 and also don't be one of the "qubic warriors" 11:24:05 qubic? Is that real thing? 11:24:08 or annoying as them 11:24:11 @hanzemito:xmr.se: "be excellent to each other and welcoming to newcomers", as per the channel topic :) 11:24:19 @octagonsupreme:matrix.org: ^^ 11:24:31 or just theory? 11:24:39 a game theory 11:24:57 #monero-offtopic:monero.social if you want to gossip about qubic 11:25:06 whatever i notice that. 11:25:21 @hanzemito:xmr.se: yes 50% of the conversation are focused on qubic instead of Monero 11:26:27 This conversation contain joking haha. 11:26:28 LOL xD 11:27:35 it seems like that p2pool is faster on my aarch64 device than monerod, at least when it comes to validating some things; i should really profile it 🤔 11:28:33 P2pool uses randomx fast mode 11:28:45 monerod does not (by defsult) 11:29:16 fast mode? 11:29:21 you mean the 2GB one? 11:29:26 Yes 11:30:24 also btw, i don't think monerod's randomX has aarch64 JIT 11:30:30 while p2pool does 11:30:42 does it? 11:30:47 No idea 11:31:27 pretty sure they both use the same implementation, but one might not be up-to-date 11:32:04 hm 11:32:13 i didn't know monerod used randomx light mode (and enforced it, seemingly), but that would explain the speed difference then 11:32:18 like how tevador's randomX has RISC-V JIT 11:32:19 (and why cuprate is faster) 11:32:21 while xmrig doesn't? 11:32:35 i mean 11:32:37 @helene:unredacted.org: It doesnt 11:32:43 some implementatios may be old :P 11:32:50 Using full mode in monerod doesnt seem to speed up sync 11:32:57 yeah, it also has aarch64 JIT, thought it might not be the most efficient > like how tevador's randomX has RISC-V JIT 11:32:59 @ofrnxmr:xmr.mx: that's odd 11:33:28 Monerod doesnt use 64bit accellerations, cuprare dkes 11:33:35 Thats one of the reasons 11:33:42 huh 11:33:56 oh so monerod is 32-bit? 11:34:04 no 11:35:25 https://github.com/monero-project/monero/pull/9856 11:36:06 > Each scalarmultKey call to the 32-bit code takes ~93 microseconds, whereas the 64-bit code takes ~37 microseconds. 11:36:25 i don't understand 11:36:34 this seems to be using the x86_64 assembly version? 11:40:45 i feel one of the main performance hits for me is how much monerod is thrashing the disks, even though i'm on NVMe drives 11:42:58 I/O scheduler adding overhead maybe? Though I can't imagine many distros default to anything other than none, or maybe kyber. 11:43:35 For NVMe drives* 11:45:46 I'm on a somewhat-custom kernel, but it's still fairly default; the possible issues I can see are the fact that they're on btrfs (CoW doesn't play well with databases, but I did disable that) and the SoC's IOMMU limitations 11:45:59 i have fairly good throughput but the IOPS/latency might suffer a fair bit 12:07:15 Good morning! What is the best wallet that can hold Monero and other crypto? (Bitcoin, mostly) 12:07:39 On what platform 12:08:04 Linux mint, and Tails. 12:08:21 Feather wallet 12:08:22 Maybe android too. 12:08:34 For monero 12:09:39 I already have that, I didn't know it did anything besides Monero. Can you give me a link to study it? 12:10:13 feather is monero-only 12:11:47 @dblmr:unredacted.org: Android.. 12:11:47 cake wallet 12:11:47 Monfluo (monero only) 12:11:47 stack wallet[... more lines follow, see https://mrelay.p2pool.observer/e/rubwsbYKNjNYSFl3 ] 12:11:56 Ok, that's what I thought. Right now all I really need is Bitcoin and Monero, but I imagine I'll need to deal with others at some point. 12:12:56 Cake, stack, and unstoppable, all support bitcoin. Cake has the most full featured monero implementation of those three 12:14:14 also has the most privacy features for other coins too, it's one of the few Bitcoin wallets to support silent payments, and one of the few Litecoin wallets to support MWEB 12:14:33 Is Cake easy enough for a crypto-challenfed old guy? 12:15:05 depends what you would usually struggle on; i feel they try to make it as user-friendly as possible but this is not an objective metric 12:18:24 I don't know enough about it to predict what problems will arise. My Linux skills are pretty good, so I can probably get it running. If Cake is a popular wallet, I imagine I'll find enough instructions on using it? 12:18:56 they have decent documentation 12:19:40 For example: https://docs.cakewallet.com/cryptos/monero/ 12:20:02 Ok, thank you very much! I guess I'll continue the day with a Cake walk! 12:20:21 Thanks for the links! 12:41:13 @ofrnxmr: A shame that none of thses are in fdroid 12:44:00 F-Droid tends to cause many issues to app developers, it's not a great platform for software in general 12:44:02 <321bob321> Gonna have fun when they only allow signed apks 12:44:34 @helene:unredacted.org: Its an issue for projects who love tracking their users, such as mozilla's stuff yeah 12:45:42 "Oh no! I cant have proprietary librairies in my GPL app!" 12:45:59 @jobotnik:matrix.org: not really, from my understanding the main problem is that it requires the projects build on F-Droid's CI infra (which keeps going down, not working and struggling) because F-Droid wants to sign all builds with their own keys 12:48:27 Cake just wants to use google's push notifications so that you have google knowing when you recieved changes on your monero wallet, big brain 12:48:49 Signal doesnt allow forks which are not using google-api 12:50:34 well, the CLI wallet isn't going away, you can still use that on your phone if you don't trust any of the options 12:50:53 @helene:unredacted.org: Why should I trust the dev for generating apks? 12:50:53 I should'nt have to trust the dev, I should only have to trust the source code. 12:50:53 The build should be reproductible, such that I dont have to trust a single-party. In case I have to trust a single party, Id rather trust a bigger org such as fdroid that a random dev 12:51:37 @helene:unredacted.org: But here its not even the central question. They dont want to use fdroid because they want to have closed source libs such as google-api 12:53:45 Closed source google stuff ? just the name makes me want to run away. 12:53:47 https://github.com/cake-tech/cake_wallet/issues/113 12:56:31 @helene:unredacted.org: Yeah, but I would advise the monero project to not promote android apps that 1. Have closed source librairies. 2. That ignore talks about such issues. 3. That give users a false sense of security by giving them an unofficial fdroid repo. 4. To have in the end google knowing when user A sent money to user B as a sidechannel 12:58:41 Is 4 known, or just suspected ? 12:59:26 From my pov, highly suspected 12:59:51 Look at signal, they dont allow forks that dont use google 13:00:08 (by suspecting, I mean the google lib has technical ability to do it and no clear above board other reason to be used) 13:01:27 Google servers have enough information, they don't have to change the lib to deanonymize the users 13:02:12 I just don't get why you would use google at all 13:02:51 Well, monero does use a (open source) google lib for tests. I guess it's useful, though honestly no idea how much it is... 13:03:16 But AFAIK it never tries to call home. Never checked though. 13:04:01 But if the lib does call google servers, it'd be a good enough red flag. 13:04:20 And that is almost certainly checkable. 13:12:39 Unstoppable wallet is > <@jobotnik:matrix.org> A shame that none of thses are in fdroid 13:13:15 Cake removed their google shit a long time ago > <@jobotnik:matrix.org> Cake just wants to use google's push notifications so that you have google knowing when you recieved changes on your monero wallet, big brain 13:28:26 @ofrnxmr:xmr.mx: I cloudn't find this change 13:28:26 There's still google gms in the gradle config 13:28:39 @jobotnik:matrix.org: Yeah, 2019? Around there 13:33:21 You're just a troll 13:33:21 mkyq a contributor of cake wallet wrote in august 2021: 13:33:21 We just back to integration into F-Droid again. We will work on option to build the app without push notifications which require google api.[... more lines follow, see https://mrelay.p2pool.observer/e/ubObtLYKanZ4TlZN ] 13:33:55 The fdroid issue was closed as a wont-fix in 2023 13:35:10 Source code is open, you can compile and use it. 13:37:08 @elongated:matrix.org: Everyone should do that? 13:37:08 The monero blockchain cannot be trusted if more than 50% of their users use proprietary apps 13:37:48 or opensource apps who's dev just want to give their users a false sense of security, and always try to go around central issues 13:37:55 @jobotnik:matrix.org: Where are you pulling these figures from ? 13:38:09 The blockchain can still be trusted as long as *you* run a daemon from good source. What you can not trust is ring degradation. 13:38:43 @jobotnik:matrix.org: It’s open source, if you can fix their issue of fdroid do a pr 13:38:46 ring signatures where every other address in the network is controlled by mkyq, yeah sounds good 13:39:27 But at worst, it devolves to better (1) than bitcoin privacy wise. [1] AFAIK, I don't know much about bitcoin. 13:40:08 @elongated:matrix.org: The community tried, however the community cannot add it to fdroid, only the main devs can. But those closed the issue as a wont-fix because they are busy 13:40:58 @jobotnik:matrix.org: Closed source issue is their exchange api part, unrelated to monero 13:43:04 Is the monero part a separate binary, where you can remove the exchange api part so it can never run ? 13:46:40 They use id "com.google.gms.google-services" version "4.3.8" 13:47:31 @jobotnik:matrix.org: They are now on v5.4.0 13:48:06 Im talking about the version of gms 13:48:12 You're clearly just a troll 13:48:23 Paid by cake wallet? 13:48:32 @jobotnik:matrix.org: 😂 13:52:28 if your security model includes google or malicious code being added to software, you should not be using anything else than the official monero CLI 14:17:53 didn't mr. **bic say he wasn't going to selfish mine anymore? 14:21:18 why would you believe anything said? 14:21:32 because it took like a day for him to back peddle and begin selfish mining... 14:21:38 *back pedal 14:22:04 so not really a question :) 14:22:08 Aaajww291: Don’t trust pubic https://x.com/c___f___b/status/1968195867770552507 14:22:37 Aaajww291: I don't recall seeing anything like that. They stopped for a marathon after the reorg and started again. 15:01:31 Hey, if my Monero tx had 11 confirmations but became unconfirmed after a re-org. Can the sender double-spend it now? Thanks. 15:11:02 @dufebo98:monero.social: Did this happen on Sunday, September 14? Were you the sender or receiver of the transaction? 15:45:57 @dufebo98: yes 15:46:25 By default, they would have to wait a week before being able to double spend it 15:47:10 Theres a pr open that would allow the double spend immediately after the initial tx was invalidated 15:57:18 @dufebo98: If there was a 21 block reorg, yes 15:57:20 ofrnxmr: isn't that risky since the initial tx is still in the txpool 15:58:13 maybe adjust the time that old transactions get flushed out from the pool 15:58:21 Cindy_: Invalidated txs should be flushed 15:58:26 ah 16:06:12 @elongated:matrix.org: eventually/ but i dont agree that double spending should go unnoticed 16:07:42 invalid are flushed after 7 days (too long imo) 16:07:42 normal tx & double spends are flushed after 3days 16:08:16 @elongated:matrix.org: No, this can happen with a12 block 16:09:15 10+n conf can be invalidated by 10+n+1 reorg 16:15:09 @ofrnxmr:xmr.mx: Wouldn’t those txs go back in mempool ? Only a rouge miner can double spend 16:16:56 No, sub-10 block reorgs go back to the mempool are are valid 16:17:59 > 10 block reorgs unconfirm txs and cause them to becone invalid. If a miner flushes the pool and wallet user re-spends these funds, it is a double spend 16:18:13 Greater than* 10 16:18:58 If minera dont flush their pools, then their nodes will reject the new tx and it will never be mined 16:19:25 Is the network doesnt flush their pools, the tx will also not be relayed (peers will reject it) 16:20:10 Well, maybe the funds went to the same initial address, so it wouldn't be a double spend. But because of stealth addresses, an external observer wouldn't know that. Only the initial intended recipient could know that. 16:21:14 Do node operators have to do anything if their node was running during the 18 block re-org? Keep getting notices of "transaction not found in pool" 16:21:43 How did all those invalid txs end up in mempool after 18 block reorg ? > <@ofrnxmr:xmr.mx> No, sub-10 block reorgs go back to the mempool are are valid 16:21:56 they were reorged with empty bloxjs 16:22:03