04:38:04 MajedticBank: yeah I can't believe how some community members behave either. Kayabas proposal breaks zero rules and has zero downvotes, yet people keep crawling out the woodworks to bitch about it. If you think the terms are unacceptable, go to gitlab and give it a downvote. 04:38:04 In the past we've merged proposals that broke rules, barely passed a 50/50 ratio on the votes (and only so with vote manipulation), had questionable value to begin with and possibly even damaged our reputation. 04:38:05 It was merged on the basis of "let the donors decide", and all I'm asking for is to give this proposal the same treatment 04:51:35 ctrwj: I don't see point on liking/disliking on each other as we are part of same community and this is not facebook, rather constructive criticism should be applied 04:52:36 opinion of one person that's true to monero and this community should be worth more then 5 likes who are here for their own reasons, not for the technology and things that monero stands for 04:52:59 let alone 3-4-5 people saying some things are not okay 05:01:54 nah bro, dislikes are the voting system. The advantage is that they are clear and leave no room for speculation. Downvote to say "don't merge in the current state" 05:03:56 Should I remind you of the time where 10+ people showed up in the event room to vote in your favor, never did anything else and then were removed by the timeout not all at once? That means they didn't even open the room to read what was happening. Dont tell me shit about "hurt durr people say the same things" when vote manipulation happens once in your favor 05:04:43 Besides, I recognize the names, and can assure you the votes on this CCS are by different people, sharing the same opinion 05:06:36 <1​23bob123:matrix.org> We are all rotten! 05:08:07 no, I'm the polite ofrn™ 05:10:23 ctrqj: Someone said in the CCS he can post his donation address and let people donate to him. I can also imagine how it would look if magic cake fund opened funding for his retroactive proposal under this terms 05:13:04 <1​23bob123:matrix.org> I’ve been accused of that too 05:20:05 Majesticbank, again we could've done the same for previous, but worse proposals. Can we please work together to make core devs their lives as easy as possible? What's the harm in putting up a CCS funding address? 05:40:01 not sure who you all are, but it looks a lot like a group of trolls who've literally been around for 5 minutes have decided to become the most active CCS participants 05:41:00 it's pretty transparent. upvoting or downvoting proposals en masse doesn't create the impression of consensus, quite the opposite 05:42:01 that's why some of us have decided to stop with the gameable metric of upvotes/downvotes and comment instead, especially on controversial topics 05:43:39 Agreed, I feel like a mix between both of them makes sense. But requiring a certain ratio of upvotes to downvotes to merge doesn't really make too much sense 06:09:42 Don't be afraid. The Monero community is so chaotic and so unruly that it's flat-out impossible to stick 100% to some rule. Nobody will be able to establish a rule "CSS will get merged only if more upvotes than downvotes" and then get people to really follow through. 06:22:19 <1​23bob123:matrix.org> no one up/down vote 06:22:33 <1​23bob123:matrix.org> just comment 07:03:36 Who has been "literally around for 5 minutes" geonic? 07:04:43 Geonic 💀 07:05:52 It's obvious that we don't blindly follow the voting ratio, and look who voted, especially when its a close call 07:06:07 A dev vote should have a higher weight on a dev proposal then eg. my vote 07:07:06 and this is how things have always been run. In the event room, more then 50% of none-majesticbank accounts voted for them, thats why they were accepted 07:07:58 The bots were pathetic ngl 07:08:16 He thought he wouldn't make it lol 07:13:29 <1​23bob123:matrix.org> you notice this comes up every couple weeks and its the same thing over and over 07:14:53 <1​23bob123:matrix.org> you make factual points and then get insulted or an alt to ofrn and then nothing happens 07:15:12 Looks like he's self reflecting. Since the people he is targeting all have been here for much longer before him. He himself only appeared after his CCS. And his proposal was upvoted en masse with some accounts accidentally sending duplicate comments supporting the proposal. 07:15:13 <1​23bob123:matrix.org> miss use email been what 3 weeks now 07:16:17 <1​23bob123:matrix.org> trying to rip off community members what 2 months since monerokon? 07:18:33 has the troll farm officially moved from -events to -community now? 07:19:02 <1​23bob123:matrix.org> thank you for making my point 07:19:13 Broken record 07:19:49 "accidentally" 07:20:14 have fun fellas 07:20:30 The accidental vote stayed 07:23:21 <1​23bob123:matrix.org> old guard need new blood i guess 07:26:14 new blood testing positive for leukemia :/ 07:27:24 let me just point out that using a CCS proposal you hate as reasoning to accept literally *anything* shows that you’re operating in bad faith 07:29:03 “what I want more for Monero is what I hate” so pls merge ok tx 07:34:46 and yeah you’ve literally been around for 5 minutes: https://repo.getmonero.org/monero-project/ccs-proposals/-/merge_requests/371#note_20457 07:35:30 btw, why are you referring to yourself in the third person here? “ct has been around..” 07:37:03 too many personalities to juggle? 07:44:16 Spend more time with people in monerokon next time. I only saw you in the lobby. Try the stands area and you'll meet us both perhaps. 07:45:23 I met him and I saw you, don’t worry. Not calling you an alt. 08:11:59 Except I have commented and voted on John Foss' Monero Moon proposal several months prior, which was deleted for some reason, so it doesn't show up in the statistics anymore 08:16:36 Regarding the repeated accusation of using alt accounts: This is the only user name I use within the monero community, CCS and everywhere else were voting matters. I do have other accounts that are used for other topics and occasionally shill monero, but that's the entire extend of involvement. 08:43:59 I'm quite disappointed with your poor memory, because we've talked about all this at least once before 09:52:43 I don't want to waste my day here, but: 09:52:43 1) I'm more receptive to feedback when it's not accompanied with a comment I should have my head bashed in 09:52:43 2) I have heard criticism. I've much less frequently heard suggestions. If anyone feels I disregarded their criticism, feel free to reformat it in a way I may do better with. Unfortunately, while I hear criticism, I'm infrequently clear on what to do in response. 09:52:44 3) Due to my acceptance the proposal may be closed, I've never seen this as a demand. I do acknowledge the retroactive nature took away the competitive market however. I believe this may be the one fault introduced to the community under retroactive funding, which has to, and be, adapted to. 09:52:44 While I did reach out to sarang for their opinion, I didn't come away with a proper answer. If the amount's validity in a competitive market is challenged, I do support the community asking CypherStack their opinion. While I don't want to ask them for a formal quote, which may take time to arrange, I would be surprised if Diego was unwilling to comment if this amount is inline/les 09:52:45 s/more than with what they'd estimate. 09:54:05 If there are actual discussions for me to participate in, and not flame wars, please ping me. The former would not be a waste :) 10:25:34 I don't want to waste my day here, but: 10:25:34 1) I'm more receptive to feedback when it's not accompanied with a comment I should have my head bashed in 10:25:35 2) I have heard criticism. I've much less frequently heard suggestions. If anyone feels I disregarded their criticism, feel free to reformat it in a way I may do better with. Unfortunately, while I hear criticism, I'm infrequently clear on what to do in response. 10:25:35 3) Due to my acceptance the proposal may be closed, I've never seen this as a demand. I do acknowledge the retroactive nature took away the competitive market however. I believe this may be the one fault introduced to the community under retroactive funding, which has to, and can be, adapted to. 10:25:36 While I did reach out to sarang for their opinion, I didn't come away with a proper answer. If the amount's validity in a competitive market is challenged, I do support the community asking CypherStack their opinion. While I don't want to ask them for a formal quote, which may take time to arrange, I would be surprised if Diego was unwilling to comment if this amount is inline/les 10:25:36 s/more than with what they'd estimate. 10:44:11 bitmain xmr miner announced price is $3k 10:47:45 so I won a coupon lol 10:48:15 you didnt like tweet subscribe follow tag retweet sorry 10:48:54 let the public tell you what they are willing to pay and pick the highest most predicted price, smart 10:49:10 https://www.asicminervalue.com/miners/bitmain/antminer-x5 10:49:10 price prediction market 10:49:46 not worth it, even with 1k discount 10:50:43 maybe just for reverse engineering fun 11:31:30 wait for 3rd hand market 11:39:10 <1​23bob123:matrix.org> Or buy second hand compacts 12:51:00 ASICS :( 12:51:01 Miner ASICS :( 12:55:52 Thanks for sharing the x5 price news ajs_ 14:49:32 You can get a 200kh/s on a 2k build ? 14:57:38 No, but you can resell your ryzen mining rig after its not efficient with mining anymore. At 2k iit might be a better deal, at 3k it is likely not 15:39:24 exactly. the response has been pathetic 15:45:23 the precedent cant be set that some one can just walk in and claim everything on their terms and disregard due process, "in the name of monero" 15:45:40 the precedent cant be set that some one can just walk in and claim everything on their terms and disregard due process, "for the benefit of monero 15:45:43 the precedent cant be set that some one can just walk in and claim everything on their terms and disregard due process, "for the benefit of monero" 16:01:03 More information about the Full Chain Membership Proofs CCS. Discussion on July 5th before the CCS proposal was posted: https://libera.monerologs.net/monero-research-lab/20230705 (FCMP is based on Curve Trees.) 16:09:21 This is the right question! xD... 16:15:00 After all my prediction of ~$3000 was correct.... Based as bitmain always does on a 1-2 year ROI without electricity discount. 16:15:01 You can't buy 218 kh in CPU + MoBo + RAM with $3k. But I guess you can't do much with an "ASIC" designed for mining only either. I mean, you can do a lot of things with 15 Ryzen 3900X besides mining. 16:15:01 Also, it would be interesting to disassemble one and see if you can somehow "reuse" those RISC-Vs 16:17:57 Rain-v should be reuseable, don’t know about the boards though 16:25:57 In Spain 13 5900x (TDP 110W) is about 4500 Euros. I do not know if there is any MoBo to install more than 1 Ryzen. Otherwise, it would be like 1000 Euros additional + Ram another 300 Euros + PSU I don't know, maybe 500 Euros if split some way. 16:25:57 So ~ 7000 USD for 218 Kh vs 3000 USD X5 16:25:58 The advantage of CPUs is re-sale and use for other tasks. The advantage of the X5 is convenience, price and footprint. 16:28:15 I see absolutely no harm in Monero. If you take the hashratede of a CPU in SHA256 vs S19j XP then the difference is abysmal. Here in Monero we are talking about how much? There is no disadvantage as there is in Bitcoin between ASIC and Home users with their computers. 16:36:27 The most important thing is that the *efficiency* is roughly on par with that of the top cpus currently available. Which means that the profitability is also the same, rather than having some massive advantage that would kill cpu mining (the way asics do in other algos) 16:39:22 You could cram 20 MH/s in a single box and sell it for $10k, but if that thing needs 20MW to run then nobody will ever buy it because it will be plain unprofitable 16:57:18 You have to understand that Bitmain's price ($3k) is not a real price that it costed them to produce these devices. They squeezed as much as they could by mining with them since 2021, and now selling for max price people are willing to pay. $3k is a price for one-time sale, not for stable production, so of course they made the price attractive. 17:11:59 Yeah! TDP is very similar 19:30:12 Does anyone see value in this? "DOCit Discord document bot" https://github.com/MAGICGrants/Monero-Fund/issues/25 19:30:28 My first reaction is to vote to reject it. 19:37:06 No, I don't feel it is worthy of funding 19:37:54 It looks like you're paying for the service and don't get the code 19:38:51 You're depending on their servers, etc. There are FOSS projects that do this already, it's just a matter of making the bot which is trivial 19:39:51 MAGIC gets lots of these types of cookie cutter proposals. I wish more worthy proposals were submitted. 19:39:57 <1​23bob123:matrix.org> Fck discord 19:40:07 I echo the same idea, excluding the expletive. But yes, discord is quite harmful 19:40:53 People want money 😄 19:40:54 <1​23bob123:matrix.org> They probably get bought out by m$ or some soon 19:45:33 idk it may be valuable 19:45:43 but not worth a lot 19:45:53 definitely reject if discord only though 19:45:54 <1​23bob123:matrix.org> They probably get bought out by m$ or someone soon 19:46:04 i dont know shit about ai but i know people that can do the same thing in a few weeks max 19:46:14 also the fact that it depends on openai's bullshit makes me not like it 19:46:24 fuck openai 19:46:38 I use this one myself: https://github.com/imartinez/privateGPT 19:46:40 No need to feed it to openai, this one is much better, just build a simple interface on top of it 19:46:50 Will take a couple weeks max 19:46:51 yep exactly 19:47:01 It can be made into a website, then send api calls to the website, or use wsgi, so many options 19:47:04 sounds like your average codemonkey that first learnt about ai when chatgpt blew up 19:47:05 <1​23bob123:matrix.org> Can i use my orb to sign in? 19:47:15 Important question ^ 19:52:17 I wonder if Monero docs are good enough to feed into the AI bot generator. And is it a good idea to depend on a large language model to help write security-critical code? 19:53:13 We just need a bot that says "Check unlock time. Check burning bug. Just use wallet2. Don't use monero-wallet-rpc over a remote connection." Done ;) 19:53:27 <1​23bob123:matrix.org> Needs to be up long enough to do it 20:47:18 kayabanerve: you're drawing a false equivalence between yourself and Cypher Stack. no one would hire CS without some sort of deliverable at the end - a report, an audit, code, something - let alone pay them for open-ended work that they initiated. 20:47:51 if you're going to demand professional rates, ask yourself what a professional relationship would look like and whether what you're doing fits the description. 20:50:38 I tried gpt4all a month ago. The level is not good for professional work. Maybe we can try to fine tuning lama2 20:52:00 But not for coding, I would say maximum for helping people to get started. Like to answer questions like how do I create a wallet, how can I run my node,how can I run p2pool 20:52:10 Defo not for coding crucial part of the project! 20:53:23 I like to abuse ai for rubber duck debugging 20:54:18 in terms of the suggestions you're asking for, this is what I suggested last week: "Kayaba can redact his proposal to include some type of milestone (literally anything that is observable and not just “I think I’m on the right track guys”), charge for the remainder of the work, whether it’s a month or a year, and it will get funded" 20:55:00 hell, even charge for the work that's already done. but include a deliverable. 20:56:41 "I've done this much work so far, I see a path towards completion, this is how many hours I need to finish it and this is the total number of hours I would've spent." 20:57:44 I think that's more than reasonable 21:00:37 I'll try llama2 soon and give a feedback here. 21:00:37 Just to understand there would be a gpu backend somewhere or even inference should run on Cpu? 21:00:37 I am experimenting a lot with bot lately so I have got already some experience. 21:00:38 The 2 main approaches here would be to: either extract context given query and use a llm as it is. Or fine tuning an existing one with docs. Time and process are obviously pretty different between the 2 choices 21:09:32 Can someone post a git link with the docs that the bot should use to answer? 21:13:56 Here a post I wrote some time ago that explains the embedding plus context plus query from user 21:13:56 https://www.the-eng-advisor.tech/how-to-use-openai-and-langchain-to-create-a-qa-bot 21:13:57 That is open AI but works fine also with any hugging face llm like lama2 21:22:21 <1​23bob123:matrix.org> Monerodoc has .md on there github 21:33:41 https://github.com/monerodocs/md/tree/master/docs 21:33:43 That is the best doc available? 21:35:25 "I wonder if Monero docs are good enough to feed into the AI bot generator. " --me, a few hours ago 21:36:52 The MAGIC Monero Fund would probably support a project to improve Monero protocol documentation. Recipients would have to KYC, unfortunately. 21:37:02 The CCS would probably support it, too.. 21:43:33 By the way the bot thing is doable. It would be nice to have it up and running in the main website. 21:43:33 BTW it is important to know if in the backend a gpu would might be available, because that does a lot of difference in the model selection