00:13:56 Is something like BCH's Flipstarter possible with Monero transactions? Inquiring minds would like to know. I have heard that Monero's script capabilities are more limited than BCH's. 00:15:08 Background reading: 00:15:08 https://www.reddit.com/r/Monero/comments/pgdi11/can_we_talk_about_some_recent_ccs_proposals/hbamnqh 00:17:58 There are no scripts in monero. 02:27:54 Ok I have more information about exactly how Flipstarter works. Quoting a BCH dev: 02:28:39 "Every coin in bch is locked and unlocked by a script but p2pkh is so standard that it’s kinda treated as not a script." 02:28:48 "To answer your question each input is a plain coin like any other. However the wallet signs a partial transaction where the output is set in stone and the signature type is anyone-can-pay so that the signature covers only the pledger’s input." 02:29:00 "Ie it’s not a p2sh. It uses a particular sighash flag. Not sure if those are available in monero"" 08:21:58 Hi, is Monero still using hardcoded IP's/Hostnames in the code for initial peer discovery ? 08:22:29 yes 08:22:46 which file ? can't find it 08:22:56 https://github.com/monero-project/monero/blob/master/src/p2p/net_node.h#L305 08:23:02 thanks! 08:23:22 there are also different seed nodes for tor/i2p 08:24:07 Are there any plans to remove them ? like in this idea: https://www.reddit.com/r/Monero/comments/f6y8xn/ccs_idea_maybe_someone_can_write_it_up/ 10:17:45 hello 10:18:24 wfaressuissia debug here http://paste.debian.net/1210366/ 10:29:24 good, reading it now 10:44:41 "https://paste.debian.net/hidden/26f00093/" recompile with it in order to see everything in gdb 10:45:16 and paste gdb backtrace for the next error 10:57:58 what is your environment (compiler, os)? 11:03:40 ok 11:09:30 debian os 11:52:20 https://paste.debian.net/1210372/ wfaressuissia 12:03:27 "https://github.com/monero-project/monero/pull/7803/files#diff-85f0577b849d28c11a899cb863aec3df1d946e0bb214a623deef05928271fef6R2548" very "smart" change 12:03:50 revert this line in 7803 patch and error will disappear 12:06:18 https://github.com/monero-project/monero/issues/7913 12:06:29 this patch? 12:08:09 it's expected that you're experiencing error only with master branch + 7803 + something else, is it correct ? 12:09:04 yes 12:09:41 "https://github.com/monero-project/monero/pull/7803/files#diff-85f0577b849d28c11a899cb863aec3df1d946e0bb214a623deef05928271fef6R2548" is it clear how to revert this line in 7803 patch ? 12:10:34 src/p2p/net_node_common.h i insert red lines and remove green 12:11:07 ok? 12:11:40 scroll a bit up in github, highlighted line is hidden 12:13:05 ah line 2548 src/p2p/net_node.inl 12:13:33 yes, only that single line 12:13:39 so all eror was a "&" !!!!!!!!! 12:13:45 no words! 12:14:32 you are capable to review that pr now 12:15:04 i i have no github account 12:15:30 create temporary one 12:21:37 "https://paste.debian.net/1210372/ wfaressuissia" before this comment It was expected that you're experiencing error with either unmodified release-v0.17 or master, but turned out it was always +7803 at least 12:22:11 please, specify always source code that you're using 12:31:56 ok 12:33:45 "well whatever im running on xmrchain.net has been up for 70 days" gingeropolous, cool 12:35:20 wfaressuissia line 2548 is with "&" ! 12:35:33 i checked now on my source code! 12:35:51 try_ping(arg.node_data, context, [peer_id_l, port_l, &context, this]() 12:36:05 and reverted one should be without '&' 12:36:41 ok! 13:13:16 wfaressuissia i confirm that revert worked 13:13:57 no more crashes 13:14:42 can u explain what that "&" did 13:17:07 it helped to summon segfault 13:21:29 I would explain, but I'm busy with something more important 13:22:07 ok thnx a lot 13:22:13 & saves a reference to the object, as opposed to copying it. If the object goes out of scope before the ref gets dereferenced, it accesses stale memory. 13:22:41 thnx moneromooo 13:28:32 moneromooo: What did you think about what the BCH dev said about sighash flags? Might Monero have these sighash flags? I am out of my depth here. 13:49:41 I don't know how bitcoin works, so I have thought about it. 13:49:53 I doubt monero has sighash flags. 13:58:49 For locking/unlocking, the only two things you can use in monero is unlock_time and double spending. 15:18:11 Hi ladies and ladiesmen, I've downloaded the full blockchain, and I want to setup 3 of my PC's to connect to my server and mine using that db, how would I do that ? 15:20:16 * Hi ladies and ladiesmen, I've downloaded the full blockchain, and I want to setup 3 of my PC's to connect to my server and mine using that db, how would I do that ? 15:20:16 Also if this is the wrong room for these type of questions, please point me in the correct direction :D 15:21:51 #monero 15:23:43 Or #monero-pools, which should know about proxies and the like you might need. 15:24:20 Or you might want to try the new p2pool instead. 15:24:37 https://p2pool.io/how 15:25:02 okok 15:25:02 ty 15:28:06 Rucknium[m]: to get the pattern where N arbitrary people fund pre-defined outputs, you would require active participation of all N people to construct the tx. 15:33:21 Oh, is it the "a single tx where several people send/sign", so the spending is atomic ? 16:15:18 moneromooo I explain it more in the reddit comment I linked above. So my understanding is that the transaction is partially signed by each participant. Once you as a contributor do your signing, no further action is required. When the last contributor tops off the budget request, a many-to-one (or many-to-many if multiple recipients are specified) transaction is broadcast. 16:15:41 This is the transaction that funded my recent Flipstarter proposal, for example: 16:15:41 https://blockchair.com/bitcoin-cash/transaction/76f089988449efb625fa12492288c54ebb758fc1d5fbe1fc35bd082eb90cd515 16:18:33 UkoeHB: So they would have to do it synchronously, not asynchronously like Flipstarter? Like schedule a specific time to get coordinate I suppose? 16:18:59 This is possible in monero, then. I have preliminary code that does that, but it's pretty complicated, and I gave up on it because it was inherently leaking the spent outputs to other signers. 16:19:25 If you do not care about this, then... you'd have to finish the code. 16:19:39 It was a PoC level only. 16:19:56 You'd get several rounds though. I think 3. 16:26:06 * specific time to coordinate 16:31:53 Rucknium[m]: it can be asynch but requires multiple rounds. If you have 20 people cooperating, then all it takes is 1 of them to block the whole process. 16:39:35 UkoeHB: Hmm interesting. With Flipstarter, contributors can cancel their pledges at any time by spending their contribution coin back to themselves. Cancellations don't block the process at all. 16:40:21 Why make it a merged tx then ? Space ? 16:43:25 It's all-or-nothing. If the fund goal isn't reached, no funds are transferred. It operationalized the idea of an assurance contract 16:43:28 https://read.cash/@flipstarter/introducing-flipstarter-695d4d50 16:53:30 "Why make it a merged tx then..." <- Only here they constantly call me using my nickname (Space)😅 16:53:57 I don't think that this is just an academic discussion about the capabilities of Monero vs BCH. There has been a lot of tension recently about "what should CCS be funding"? Part of that tension derives from the somewhat permissioned nature of CCS. 16:55:20 Well if the technical capabilities aren’t there it’s kinda a dead end, unless you can come up with a solid technical solution. 16:55:22 CCS literally stands for Community Crowdfunding System. it should fund whatever the community wants it to fund. 16:56:49 UkoeHB: Right. So is the final word what you said about one person being able to block the process, so it wouldn't stand up to adversarial attack? 16:56:54 The recent tension you refer to is just noise. 16:59:04 hyc: To me, it is clearly not just noise. It caused Monerujo to abandon their CCS proposal and try their own way of requesting contributions: 16:59:04 https://np.reddit.com/r/Monerujo/comments/pe0g89/call_for_donations_monerujo_sidekick/: 17:07:46 Ah I see. It should be possible in monero as well. Every person signs their ring(s). 17:09:25 :D 17:09:41 moneromooo: Thank you for looking into it. 17:10:00 To be clear, I'm not looking into it. 17:10:14 I have in the past, but I have no interest it dusting it back into shape, 17:10:53 I mean, you looked into it by checking to see if it would be possible. I didn't mean to imply that you would do further thinking or looking 17:10:59 I can find the code if someone else wants to though and want to reuse it. 17:25:17 I don't think "a lot of tension" is a fair characterization at all. A project decided to go its own way for funding. 17:28:11 There is this 17:28:11 https://np.reddit.com/r/Monero/comments/pgdi11/can_we_talk_about_some_recent_ccs_proposals/ 17:28:40 And in the recent past this 17:28:40 https://repo.getmonero.org/monero-project/ccs-proposals/-/merge_requests/222 17:36:14 that latter one is, IMO, more about for-profit corporations soliciting community donations. which is in poor taste IMO 17:37:04 if you're a for-profit corp and need funding to get off the ground, give each backer shares of the company. that's capitalism. 17:37:27 donation-based funding should only be for nonprofit efforts. 17:42:20 This is exactly what I am referring to. With a permissionless, self-hosted, noncustodial approach like Flipstarter, there would really be no questions about what "the community" should fund. Flipstarter is just individuals autonomously cooperating to ensure that capital meets labor and produces something useful to someone. 17:43:43 moneromooo: the current protocol is not well-suited to Rucknium’s usecase because it is not resistant to attackers. All inputs sign the same msg (all tx data) which means all inputs must be known before you sign. A malicious participant can block other signers by just not signing. I think it is feasible to change this at the protocol level, so CLSAGs only sign their own content and the tx outputs. This way inputs are 17:43:43 independent of each other. 17:44:27 Then, if there is a malicious individual, the tx author can just find another person to finish funding the tx. 17:44:53 It would not at all be a replacement for the existing CCS. The CCS definitely has certain advantages over Flipstarter. For instance, with CCS there is a defined process for proposal feedback before asking for funds. Ensuring accountability from the proposer is also more difficult since the funds are released immediately upon the fund target being reached. 17:45:25 lack of accountability sounds like a non-starter 17:47:16 Flipstarter has funded 79 projects with over 9000 BCH. It has been working reasonably well for BCH. Just an idea. I am trying to see inefficiencies and overcome them if possible. 17:48:38 For Flipstarter's difficulties with accountability, see 17:48:38 https://np.reddit.com/r/btc/comments/ofru2c/what_happened_flipstarter_campaigns_that_did_not/ 17:50:13 philosophically it'd be like zero-conf txns: I'd trust it only with small amounts that i don't care about losing 17:51:44 The accountability problem can be mitigated by setting up a repeated game: Do funding in stages. That's what my proposal did: 17:51:44 https://flipstarter.redteam.cash/ 17:52:22 Proposers having reputations also helps mitigate the accountability issue. 20:25:46 Solar freakin' roadways! 20:31:44 I mean, do what you want. I am not the right person to champion this nor implement it. I am just trying to determine if it is possible and laying out a case for it. 20:33:56 (congrats on getting fully funded!) 20:36:08 plowsof: Thanks! It was funded within just 48 hours of going live. I'm going to submit a CCS soon too, to work on the mixin selection algorithm with jberman. Then I will write a comparative review of CCS and Flipstarter :) 22:01:50 how to change block reward not fees only block reward 22:04:39 marasli: can't be changed 22:05:02 why 22:05:43 Them's the breaks 22:06:06 for fork not monero 22:06:39 Gotta go look at the code I guess ^.^ 22:07:20 i'm not sure, i checked 22:10:40 "DYNAMIC_FEE_PER_KB_BASE_BLOCK_REWARD must be divisible by 1000000" why? 23:07:20 are there any people here that use `make release-static` regularly instead of `make depends`? 23:07:49 https://github.com/monero-project/monero/pull/7773 might make `make release-static` more complicated because static unbound has to be compiled manually 23:08:18 but I don't know how much we should care about it 23:09:25 I do. Though I usually monkey with the cmake line for it. 23:12:15 I wouldn't be overly annoyed by having to build unbound separately though.