14:00:29 <achow101> #startmeeting 14:00:41 <achow101> #bitcoin -core-dev Meeting: achow101 _aj_ amiti ariard aureleoules b10c BlueMatt brunoerg cfields darosior dergoegge dongcarl fanquake fjahr furszy gleb glozow hebasto instagibbs jamesob jarolrod jonatack josibake kallewoof kanzure kouloumos kvaciral laanwj LarryRuane lightlike luke-jr MacroFake Murch phantomcircuit pinheadmz promag provoostenator ryanofsky sdaftuar S3RK stickies-v sipa theStack TheCharlatan vasild 14:00:46 <hebasto> hi 14:00:47 <josie> hi 14:00:48 <TheCharlatan> hi 14:00:49 <virtu> hi 14:00:51 <glozow> hi 14:00:52 <sipa> hi 14:00:52 <brunoerg> hi 14:00:53 <ajonas> Hi 14:00:55 <Chris_Stewart_5> hi 14:00:57 <pinheadmz> hi 14:00:59 <laanwj> hi 14:01:01 <sdaftuar> oh hi 14:01:09 <vasild> hi 14:01:10 <achow101> There are 2 preproposed meeting topics this week. Any last minute ones to add? 14:01:12 <willcl-ark> Hi 14:01:12 <stickies-v> hi 14:01:15 <theStack> hi 14:01:43 <achow101> #topic package relay updates (glozow) 14:01:45 <Murch[m]> Hi 14:01:51 <pinheadmz> oh, hi mark 14:01:55 <glozow> #28970 was merged (yay!) and a followup is open: #30012 14:01:55 <glozow> #30000 is the next PR in the "p2p track" though relatively small. I'm currently working on addressing some feedback on it. The next one will be TxDownloadManager which I am almost done rebasing. 14:01:59 <gribble> https://github.com/bitcoin/bitcoin/issues/28970 | p2p: opportunistically accept 1-parent-1-child packages by glozow ÷ Pull Request #28970 ÷ bitcoin/bitcoin ÷ GitHub 14:02:00 <gribble> https://github.com/bitcoin/bitcoin/issues/30012 | opportunistic 1p1c followups by glozow ÷ Pull Request #30012 ÷ bitcoin/bitcoin ÷ GitHub 14:02:01 <glozow> The priority PR is #28984 14:02:02 <gribble> https://github.com/bitcoin/bitcoin/issues/30000 | p2p: index TxOrphanage by wtxid, allow entries with same txid by glozow ÷ Pull Request #30000 ÷ bitcoin/bitcoin ÷ GitHub 14:02:04 <lightlike> Hi 14:02:04 <gribble> https://github.com/bitcoin/bitcoin/issues/28984 | Cluster size 2 package rbf by instagibbs ÷ Pull Request #28984 ÷ bitcoin/bitcoin ÷ GitHub 14:02:06 <abubakarsadiq> hi 14:02:14 <glozow> (that should go in the high prio for review board) 14:02:38 <glozow> instagibbs posted yesterday about a WG to review the PR together, for those who missed the message 14:02:48 <instagibbs> ð 14:03:31 <achow101> added 28984 to the board 14:03:31 <glozow> That's all from me 14:03:35 <glozow> achow101: thanks! 14:03:48 <achow101> #topic cluster mempool updates (sdaftuar) 14:04:23 <sdaftuar> since last week, i was able to rebase PR #28676, which i did and then eventually got CI happy (nice tests everyone). since then, it needs to be rebased again, which i'm working on 14:04:26 <gribble> https://github.com/bitcoin/bitcoin/issues/28676 | [WIP] Cluster mempool implementation by sdaftuar ÷ Pull Request #28676 ÷ bitcoin/bitcoin ÷ GitHub 14:05:50 <sdaftuar> i was also able to get a draft reimplementation of mini-miner that makes sense post-cluster mempool -- figuring that out was one of my todo's but i don't think it will need to be part of the initial cluster mempool PR 14:06:07 <sdaftuar> hwoever, happy to share that with anyone who is interested (and can open a draft PR that builds on #28676) 14:06:09 <gribble> https://github.com/bitcoin/bitcoin/issues/28676 | [WIP] Cluster mempool implementation by sdaftuar ÷ Pull Request #28676 ÷ bitcoin/bitcoin ÷ GitHub 14:06:32 <sdaftuar> i also continue to poke sipa with a stick to get him to PR his cluster linearization code :) 14:06:35 <sdaftuar> that's it from me! 14:07:01 <sipa> as for my work, i'm back on working on a PR for low-level linearization code after figuring out enough about the idea i mentioned last week 14:07:27 <sipa> which will be not integrated into anything, but well tested on its own 14:07:58 <glozow> nice! 14:08:00 <instagibbs> looking forward to it, ideally much more testable 14:08:05 <b10c> hi 14:09:11 <achow101> #topic legacy wallet removal updates (achow101) 14:09:16 <dergoegge> hi 14:09:44 <achow101> hasn't been much activity on this project, waiting for more review on #26606 14:09:46 <gribble> https://github.com/bitcoin/bitcoin/issues/26606 | wallet: Implement independent BDB parser by achow101 ÷ Pull Request #26606 ÷ bitcoin/bitcoin ÷ GitHub 14:10:02 <fjahr> hi 14:10:34 <josie> achow101: on my list to review, feel free to pester me if you dont get a review before next meeting 14:10:47 <achow101> josie: will pester :) 14:11:00 <achow101> #topic Ad-hoc high priority for review 14:11:02 <achow101> Anything to add or remove from https://github.com/orgs/bitcoin/projects/1/views/4 14:11:09 <dergoegge> (sipa: just in case you're not aware, CI failed on #29625) 14:11:11 <gribble> https://github.com/bitcoin/bitcoin/issues/29625 | Several randomness improvements by sipa ÷ Pull Request #29625 ÷ bitcoin/bitcoin ÷ GitHub 14:11:16 <sipa> dergoegge: oh, thanks! 14:12:40 <achow101> #topic moderation guidelines (achow101) 14:13:24 <achow101> ajonas wrote some moderation guidelines as a place for us to start thinking about this topic. I've put them into a gist, if anyone has feedback: https://gist.github.com/achow101/9192ad26dc4ef08e9c899caeddc968ef 14:14:39 <pinheadmz> for a test, i uploaded these guidelines to GPT and instructed it to evaluate some comments with either "OK" or "needs moderation" then fed it some comments from the datacarrier PR. The results were amazing, only one false positive 14:15:09 <darosior> pinheadmz: hah, neat 14:15:11 <glozow> pinheadmz: woah cool 14:15:26 <pinheadmz> i already have webhooks from the repo for telegram and IRC bots, i could send all comments to the bot to at least help flag incidents 14:15:32 <achow101> pinheadmz: you mean you agreed with all of it's decisions except for one? 14:15:45 <pinheadmz> https://chat.openai.com/share/3595a095-5918-4200-91e5-97d536899e51 14:16:00 <pinheadmz> achow101 correct, it flagged a long comment by Murch[m] as "needs moderation" which i disagreed with 14:16:01 <sipa> pinheadmz: does that mean that "you" are volunteering for the role of moderator? 14:16:07 <pinheadmz> sipa yes 14:16:51 <josie> ajonas: looks great on first read, only one that jumps out to me as potentially hard to reason about is "contains false statements" 14:17:01 <ajonas> i think it'd be healthy to have more than 1 volunteer 14:17:16 <glozow> can we volunteer other people? 14:17:17 <luke-jr> seems like some moderation outcomes could require context 14:17:50 <pinheadmz> luke-jr of course and just to be clear, I would only use GPT to get my attention, then use my brain to make decisions 14:18:30 <achow101> if you have comments on the guidelines, please leave them as a comment on the gist 14:18:45 <glozow> fwiw I'm in favor of pinheadmz being one of the moderators! 14:19:16 <achow101> i'd like for us to discuss who to add as moderators in a week or two after the guidelines have had some review 14:19:22 <josie> +1 on pinheadmz, they've already been operating in a similar role with issue triage (and doing a fantastic job) 14:19:39 <glozow> my bad! sounds good 14:20:05 <willcl-ark> I'd be happy to volunteer myself to help with that too if helpful 14:20:14 <pinheadmz> +1 for willcl-ark 14:20:35 <hebasto> +1 for willcl-ark 14:20:35 <ajonas> josie: yes, I think it's directionally correct but suggestions welcome 14:21:29 <josie> +1 for willcl-ark (who is also doing a great job in the issue triaging role, which feels like it has some overlap with moderation) 14:21:54 <stickies-v> +1 for pinheadmz and willcl-ark , thank you 14:22:03 <achow101> #topic Knots use of the Bitcoin Core project on Transifex.com (hebasto) 14:22:10 <hebasto> hi 14:22:14 <hebasto> To translate the GUI, we use Transifex, both a website and command-line tools. 14:22:25 <hebasto> The history of recent events is as follow: 14:22:32 <hebasto> 2024-03-24 -- a new file 'knots-translation-26x' was added to the project bitcoin 14:22:37 <hebasto> There were no announcements or any preliminary discussions with people who are related to GUI translation. 14:22:45 <hebasto> I asked luke-jr on IRC the same day. His response was "yes, there's been some interest in translating the strings missing in Core too; bonus this could mean when those things get merged into Core, there's already a translation ready to go" 14:22:52 <hebasto> 2024-04-26 The file 'Bitcoin Knots 26.x' in project 'bitcoin' was updated with new content. 14:23:00 <hebasto> People - Bitcoin Core translators - on Transifex started to complain: "If the bitcoin-knots ressources are supposed to belong the a forked project of bitcoin-core (bitcoin-knots), i would prefer a new project being setup here on Transifex, since its kinda hijacking the main bitcoin core project and binding translation effort from members who want to support the bitcoin core project from my impression." 14:23:07 <hebasto> I asked luke-jr on IRC to remove Knots resources from Transifex. Still no response from him. 14:23:13 <hebasto> That's it. 14:23:19 <hebasto> The question is should we allow to use the Bitcoin Core on Transifex.com by any other project? 14:23:51 <achow101> I agree that the knots translation should be separate from bitcoin core's 14:24:00 <sipa> I don't see why the Bitcoin Core transifex project should cater to translations of other projects. 14:24:11 <luke-jr> I received no such request, and the complaints sound like trolling 14:24:23 <luke-jr> and the transifex project is not exclusively Bitcoin Core's to begin with 14:24:38 <achow101> I don't think it makes sense for us to support forks, even though there is likely to be significant overlap 14:25:09 <achow101> luke-jr: how so? 14:25:13 <dergoegge> achow101: +1 14:25:18 <laanwj> i agree, please split off the transifex for different software, it's confusing for users and for us 14:25:34 <luke-jr> It's the same software, just different releases 14:26:03 <laanwj> if it's the same software you can just use our translations, no need to add a project :) 14:26:54 <luke-jr> it's just another resource, like any other release 14:26:57 <stickies-v> +1 for splitting off, our project should be just for Bitcoin Core, not forks 14:27:15 <luke-jr> Bitcoin Knots is part of the Bitcoin Core project anyway 14:27:20 <dergoegge> lol 14:27:22 <achow101> no it isn't 14:27:28 <josie> luke-jr: not its not? 14:27:39 <achow101> we don't make releases for it. we don't review prs for it. it's not listed on bitcoincore.org 14:27:47 <luke-jr> yes we do 14:28:04 <sipa> luke-jr: *you* do, because it is your project 14:28:13 <lightlike> +1 to removing from transifex 14:28:33 <sdaftuar> why do we need luke to do anything? how do permissions work with transifex? 14:28:51 <glozow> +1, I don't think we should do translations for Bitcoin Knots within Bitcoin Core transifex 14:28:59 <darosior> blinks - thinks "no this is indeed real" 14:29:06 <hebasto> there are 4 admins: laanwj hebasto luke-jr and seone 14:29:06 <luke-jr> if you want to be trolls, you can make your own transifex project instead of trying to hijack this one; I didn't give laanwj access for that reason 14:29:33 <achow101> luke-jr: how is this trolling? 14:29:39 <sdaftuar> hebasto: ah thank you for explaining 14:29:42 <luke-jr> achow101: there is zero reason to have separate projects 14:29:50 <achow101> I don't think anyone else has ever had the understanding that knots and core are part of the same project 14:29:55 <luke-jr> both Core and Knots benefit from using the same one 14:30:37 <laanwj> it's just because the history of the transifex project is somewhat weird, it used to be owned by a person. we don't even know, i think Warren Togami transferred it to me at some point 14:31:01 <laanwj> (well not him, he knows someone there) 14:31:21 <luke-jr> no, I gave you access 14:32:28 <sdaftuar> luke-jr: frankly i think your position is absurd 14:32:31 <hebasto> ah, and wtogami is also a transifex maintainer 14:32:40 <sipa> I think it's been abundantly clear to anyone that the bitcoin project on transfex is bitcoin core's translation - it mentions that software in the description, and links to the bitcoincore.org website 14:33:13 <sipa> And I don't see the benefit to Bitcoin Core of having the Knots translations be part of the same project. 14:33:19 <achow101> the benefit is only because knots merges in prs before they've been reviewed or finished, and includes several things that are unlikely to be merged into core. the benefit to knots is of course that you can reuse our translation files, but you should already be able to do that anyways 14:33:36 <pinheadmz> what do the bitcoin forks use ? 14:33:54 <luke-jr> sipa: having translations ready in advance of PRs being merged into Core 14:34:11 <achow101> this seems like it would add an additional burden on translators as they would be prompted to translate strings that won't show up in Core, or may change in the future because the pr is unfinished, or have to translate multiple versions of essentially the same string 14:34:18 <sdaftuar> luke-jr: presumably that also wastes translator time when translation needs changed during PR review 14:34:44 <achow101> pinheadmz: their own transifex if they even bother with translations 14:34:54 <achow101> (or some other translation platform entirely) 14:35:13 <sipa> luke-jr: fair enough, that's a tiny advantage; but it also means wasting translator's time on changes that don't end up in Bitcoin Core in the same form, or never at all 14:35:58 <darosior> people who are actually doing the work have complained about this luke-jr, why should you insist on putting the burden on them for your own project while they explicitly don't consent? 14:36:16 <luke-jr> darosior: they're free to not do it, it's clearly marked as Knots 14:36:20 <sipa> Bitcoin Knots is clearly a different project, with different maintainer(s), different focus, different code; it's not Bitcoin Core's contributors/translators job to assist with that. 14:36:31 <hebasto> darosior: not about luke-jr but about knots resources to translate 14:36:39 <luke-jr> darosior: I haven't seen any such complaints, but I doubt it's many compared to number of people involved 14:37:04 <bitcoin-git> [bitcoin] jasonandjay opened pull request #30023: doc: add dustThreshold explain of P2SH (master...dustThreshold-P2SH) https://github.com/bitcoin/bitcoin/pull/30023 14:37:10 <luke-jr> sipa: it's produced with the same development process/contributors as Core 14:37:20 <darosior> It's these complaints which sparked this discussion in the first place 14:37:21 <sipa> luke-jr: lol 14:37:25 <luke-jr> and Transifex/bitcoin is not Bitcoin Core's exclusively 14:37:26 <achow101> hebasto: were these complaints made in private? It doesn't seem like transifex has some kind of discussion area 14:37:36 <sdaftuar> i don't think the legitimacy of our project should be used to lend credibility to Bitcoin Knots 14:37:45 <sipa> sdaftuar: +1 14:38:01 <Earnestly> darosior: Do you have any links to them? 14:38:05 <hebasto> transifex has internal messaging sysytem 14:38:12 <darosior> Earnestly: cf hebasto's comment above 14:38:18 <Earnestly> Oh, so they're not public? 14:38:23 <josie> luke-jr: you taking the position that core and knots are the same project is odd considering publicly ive seen you refer to them as separate projects 14:38:34 <darosior> sdaftuar: +1 14:38:35 <hebasto> Earnestly: no 14:38:44 <achow101> luke-jr: when I go to the project, the description says "Bitcoin Core". The homepage link goes to bitcoincore.org. I don't see how that is not indicative that it is exclusively for Bitcoin Core 14:39:00 <achow101> the translation instructions go to our repo 14:39:09 <laanwj> yes... 14:39:29 <Earnestly> darosior: Mostly curious because this was prompted not by complaints but by hebasto noticing "a new file 'knots-translation-26x'" without mentioning complaints 14:39:31 <vasild> Well, I think that at this point it is obvious that luke-jr is at odds with others, will not change his mind and will not cooperate. 14:40:05 <darosior> vasild: +1 this is not going anywhere 14:40:23 <Earnestly> vasild: To be fair, what cooporation has been offered? 14:40:28 <sdaftuar> based on this, i don't think luke should have admin rights over any aspect of our project, including the transifex site. i will let others weigh in on how best to achieve that. 14:40:30 <brunoerg> vasild: +1 14:41:11 <achow101> In any case, it seems like everyone here disagrees with luke and think the resource should be removed, so we should just do that 14:41:31 <vasild> Earnestly: by cooperate, I mean to remove the knots file from the translation site (I assume he added it there) 14:41:32 <Earnestly> So much for coorporation; do it our way or we'll just remove you 14:41:33 <luke-jr> sdaftuar: you are now attempting a hostile takeover of the Transifex project 14:41:38 <dergoegge> achow101: +1 14:41:41 <sdaftuar> sdaftuar: i have no admin rights there 14:41:55 <sdaftuar> (oops, i'm talkin gto myself!) 14:41:57 <laanwj> sdaftuari agree, luke-jr doesn't seem reasonable about it, even if everyone in the project wants it to change 14:42:19 <sdaftuar> luke-jr: i have no authority to remove you myself, but yes i'm expressing support to remove you, or suggesting you step down 14:42:26 <achow101> Earnestly: I don't understand what you mean by "cooperate". Either it's there or it isn't, or do you have some third alternative in mind? 14:42:43 <Earnestly> achow101: Just leaving it alone, there's no evidence of harm 14:42:53 <achow101> also, in hebasto's opening statement "People - Bitcoin Core translators - on Transifex started to complain" 14:42:58 <Earnestly> Or create a new transifex project for bitcoin core specifically 14:43:12 <Earnestly> achow101: Yes, but it would be nice to have evidence 14:43:12 <sipa> Earnestly: we have one; that's the one we are talking about, obviously 14:43:43 <sipa> it's ludicrous to suddenly assert that the transifex is also for Knots, or even that Core and Knots are the same project altogether. 14:43:49 <Earnestly> sipa: It's just called "bitcoin", I was suggesting creating a "bitcoin core" project 14:43:52 <achow101> Earnestly: other than luke's assertions, there's no indication that this transifex project is not Bitcoin Core's exclusively 14:44:03 <sipa> Earnestly: https://github.com/bitcoin/bitcoin/ 14:44:06 <josie> Earnestly: since knots was added recently (for the first time?) seems much easier to create a knots specific transifix project 14:44:22 <vasild> Earnestly: github's repo is also only "bitcoin": github.com/bitcoin/bitcoin 14:44:27 <sipa> Earnestly: for historic reasons, that is the URL for the Bitcoin Core github project, and the same holds for Transfix 14:44:58 <stickies-v> +1 on removing the knots resource and luke-jr's admin rights if that's possible, this conversation is absurd 14:44:58 <sipa> *Transifex 14:45:02 <Earnestly> vasild: Sure, but it's called bitcoin core now, that was its historical name. It appears the transifex project similarly has a longer history than bitcoin core as it has come to be 14:46:13 <josie> Earnestly: knots aside, are you advocating that the transifex project is a catch all for *any* bitcoin software (e.g. btcd) 14:46:43 <Earnestly> josie: No I think it should have its own project, operated and controlled by the same people 14:46:48 <sipa> let's please not get further in the rabbithole of discussing URLs, that's really a separate discussion 14:47:03 <Earnestly> josie: Which would avoid this entire situation 14:47:26 <achow101> anyways, I think we've answered hebasto's question. 14:47:30 <achow101> Any other topics to discuss? 14:47:37 <vasild> luke-jr: "sdaftuar: you are now attempting a hostile takeover of the Transifex project" -- when people cannot reach an agreement, what other options are there? The current situation can be viewed as a hostile hijack of the bitcoin core's translation by bitcoin knots 14:48:14 <Earnestly> vasild: (To be fair, this channel has been nothing but hostile to luke-jr for as long as I've seen it; perhaps warrented, but also means constantly defensive) 14:48:45 <darosior> Earnestly: that's not fair. 14:48:50 <instagibbs> I've seen no assumptions of bad faith except by luke-jr 𤷠14:49:05 <luke-jr> josie: especially if they share most strings, I don't see why not 14:49:19 <achow101> #endmeeting