Time Nick Message 03:39 MTDiscord It looks like for the MT mapgen settings that end with , eased that flag actually isn't saved, and it shows up as defaults in map_meta.txt 03:40 MTDiscord e.g mgflat_np_terrain, which is noise_params_2d 0, 1, (600, 600, 600), 7244, 5, 0.6, 2.0, eased by default 06:12 celeron55 Zughy[m]: i don't think anyone would have anything against you leading a meeting. it's just bringing up and closing the meeting topics so not a big deal regardless of how you think about it 06:21 celeron55 and the meeting topics generally being issues and PRs it's something you definitely know something about 08:17 nrz i was afk previous evening, about formspec, having rendering server side show the issue we have with formspec size . We don't have relative notation in order to size the formspec ? for example: 80% of the parent like in CSS ? I didn't coded a formspec for a while or see any information about a such feature 08:19 nrz that prevent complex things like storing end user window size server side 08:19 nrz (and also it shows that the limits of our server side formspec, and we need at a point to have real client side formspecs with actions server side) 08:36 sfan5 merging #12695 #12652 in 10m 08:36 ShadowBot https://github.com/minetest/minetest/issues/12695 -- Cut back on Gitlab-ci & misc. updates by sfan5 08:36 ShadowBot https://github.com/minetest/minetest/issues/12652 -- Remove resolution of appstream screenshots by farribeiro 08:39 lionkor its really fun to work on this, its a wonderful workflow 08:42 rubenwardy #12703 08:42 ShadowBot https://github.com/minetest/minetest/issues/12703 -- ContentDB: Fix crash when trying to overwrite a package by rubenwardy 08:42 rubenwardy also adds unit tests to pkgmgr.install_dir 08:43 rubenwardy found a few other bugs too 09:37 Zughy[m] fun fact: 117 issues are by Wuzzy https://github.com/minetest/minetest/issues/created_by/wuzzy2 09:43 sfan5 looks like gitlab moved repository mirroring to the paid plan which is why half our repos don't mirror anymore 09:43 sfan5 what a joke 09:43 sfan5 and this is what people have been constantly suggesting we move to? 09:57 Zughy[m] People have been suggesting something that doesn't belong to some big corporation, waiting to be sabotaged like they did with a lot of other products (see Minecraft for instance, or Atom, to name recent cases) 09:58 Zughy[m] nobody has ever said that GL is perfect 10:17 rubenwardy GitLab is self-sabotaging faster than Microsoft can ruin GitHub 10:22 rubenwardy GitLab wasn't really a good alternative as it's still owned by a company and isn't fully open source 10:27 sfan5 I can understand that gitlab has to use free handouts more sparingly but for a FOSS project really moving to a worse platform is out of question 10:41 Zughy[m] then you have people suggesting either Codeberg or Gitea, but I guess this is more of a meeting topic 11:13 nrz sfan5: honestly ? Maybe we need to do some déclaration about Oss project... 11:13 nrz I like gitlab but their governance on their public instance is Not on their favor 11:25 Zughy[m] the point is, if we ever change, a lot of active contributors will follow no matter where we go. If sfan, krock, ruben, x2048 and lhofhansl move, you already have all the active core devs. Then there would be me (triager), MisterE (blog) and long time contributors such as Wuzzy and - I'm pretty sure - appguru. Congratulations, you still have Minetest 11:25 Zughy[m] moreover, we'll be able to set a good example and fuel the development of the platform we've chosen 11:34 MTDiscord What would be the issue with a full mirroring? 11:36 lionkor I've only contributed once but anything that takes my SSH and GPG key and has a PR UI is good enough 11:37 lionkor you might wanna split docs like the lua api doc on github to keep the barrier low for it 11:39 Zughy[m] related to docs: this seems dead https://github.com/minetest/minetest_docs 11:40 MTDiscord It unfortunately pretty much is 11:41 MTDiscord I have a couple large-ish WIP PRs 11:43 Zughy[m] I think MT should make up its mind about documentation: it's everywhere, and usually not up-to-date (see the dev wiki) 11:43 MTDiscord Agreed 11:43 MTDiscord The problem just is that we're lacking manpower everywhere ? 11:44 MTDiscord Especially for the chore-ish tasks like mt_docs 11:45 Zughy[m] I think that the real problem is, there is no real assessment for these things and they happen "just because". And that's what you get 11:48 Zughy[m] when you (plural) decided to bring that project to life, did both the parties (you and who said yes) consider the manpower needed in the first place and the average free time you spend on MT? 11:50 Zughy[m] And, out of curiosity, why was this agreed upon without dismantling the Lua API section of the dev wiki first (which makes more harm than good)? Sorry, this is my management side speaking 11:52 MTDiscord tightens tie 11:52 MTDiscord Because I have no power there 11:52 MTDiscord I can't edit the dev wiki 11:52 MTDiscord Pretty much only the devs (and Wuzzy?) can 11:52 lionkor doesn't that defeat the point of a wiki :D 11:54 rubenwardy anyone can request access to the dev wiki 11:54 rubenwardy or player wiki 11:54 MTDiscord Zughy: To answer your question: No, we certainly didn't consider time management 11:54 MTDiscord Otherwise we would find time for it 11:55 Zughy[m] well, as a modder and core team member, you have my blessing to dismantle that section. Pointing to the modding book and the lua_api.txt is enough. I don't know what the others think though 11:58 Zughy[m] RE: "no we certainly didn't consider time management" 11:58 Zughy[m] Not good, at all. Basically now we have something unfinished taking time from one of the most active engine contributors, potentially stressing them out 12:00 Zughy[m] Also because it's not a one time thing. Every time the engine gets some major updates, that documentation must be changed too. And we both know this is not going to happen 12:03 Zughy[m] IMO people can already help ruben with the modding book if they want, even if we're talking about two slightly different things. There's no need to increase the pile of things we should care about, when we clearly don't have enough manpower/time. If core devs would ever consider to dismantle the minetest_docs repo as well, you have my +1 12:04 sfan5 as long as I'm not required to update that repo all is fine by me 12:32 MTDiscord "anyone can request access to the dev wiki" that access seems to just allow for plastering 90% of the dev wiki with "UNOFFICIAL DO NOT USE" rather than actually deleting the pages and doing the cleanup, in the case of wuzzy's access to it 12:35 MTDiscord Is deleting the pages permanent or can they be brought back? Either way make sure to back up pages. 12:38 Zughy[m] Why do you need a backup of something broken? 12:40 MTDiscord the cleanup would mean deleting things that are completely redundant and outdated, mediawiki has a page export feature and I think also keep revision history for deleted pages but in general what would be deleted would be things that don't serve any purpose to anyone to warrant a backup 12:54 celeron55 does anyone have an idea about how many % of the dev wiki actually is useful 12:56 celeron55 certainly only a few pages are useful to me - those related to the organization 12:57 MTDiscord well the amount of pages proposed for deletion is 314 and the amount of content pages (i.e. main article pages, not userpages or redirects) is 356 so... about 13% of it 12:58 celeron55 would a forum or discord poll make sense to figure out whether newcomers find it more useful than us? 12:59 celeron55 my proposal is to just delete the pages proposed for deletion, and label the wiki to be intended more towards organizational things and engine internals rather than API documentation 13:01 celeron55 replacing the content of pages with links to elsewhere about relevant subjects probably makes sense 13:01 celeron55 it wouldn't be nice to end up in the wiki and not find any help at all 13:02 Zughy[m] agreed 13:02 MTDiscord yeah pretty much my thoughts exactly 13:03 MTDiscord I do find the dev wiki to have some useful gems relating to the engine, but having it also contain lua API documentation was a bad idea 13:03 Zughy[m] actually, we can't take care not even of the engine. Proof: https://github.com/minetest/minetest/issues/10056 13:05 MTDiscord most newcomers probably only know the dev wiki as the site with odd api documentation that people in the discord server tells you not to use, if that was gone then the actual HTML rendered pages of lua_api.txt would probably appear above it in google search results 13:07 celeron55 Zughy[m]: i suspect most don't even realize there is some engine documentation in the devwiki 13:08 Zughy[m] probably, but the point is no one has taken care of it in these years, except when someone else had pushed to actually write something (and then it was dead again) 13:08 Zughy[m] which I understand, considering the amount of work core devs already have to face. At least, PRs are decreasing now 13:09 celeron55 well the thing is, even if it's old it's not irrelevant. MT's structure doesn't change that much, just some bells and whistles are added now and then 13:12 celeron55 things have been reworked over time but rework generally doesn't change the structure or how it works, it just makes it a bit neater in one way or another 13:33 Zughy[m] so we keep the engine and organisation sections, blowing up the API part? It sounds great to me 13:33 Zughy[m] Actually, can I have an account as well? To edit the meeting section when I need to 13:33 Zughy[m] I had one but it doesn't work anymore 13:36 rubenwardy PM me your email 13:38 rubenwardy oh Zughy already exists, maybe celeron55 can reset your password? 13:38 Zughy[m] let me see if I can reset it 13:38 rubenwardy found the form 13:38 rubenwardy lets see if wiki emails work now 13:38 rubenwardy x doubt 13:39 celeron55 good luck 13:40 celeron55 i think i managed to make the forum send emails reasonably well but the wikis are still a mystery 13:41 celeron55 think of it as extra security: an attacker can't attempt a password reset! 13:41 Zughy[m] done, I'm in 13:42 celeron55 ok so there's no extra security, what a bummer 13:56 Zughy[m] I can't delete wiki pages, apparently I need some high level permission: https://www.mediawiki.org/wiki/Help_talk:Deleting_a_page 13:58 celeron55 are there many pages that should be just deleted without leaving a link to somewhere else? 13:59 Zughy[m] I guess all the Lua API pages 13:59 Zughy[m] lel https://dev.minetest.net/TODO#TODO_in_Minetest_Game 14:00 sfan5 now that amkes me wonder what are terasologys trees like? 14:00 celeron55 i'd consider emptying the pages but adding a link to the relevant section in https://minetest.gitlab.io/minetest/ or something 14:01 sfan5 makes* 14:02 celeron55 would also give some SEO to the linked better documentation page 14:02 celeron55 i mean, it's stupid but it works 14:02 celeron55 so it's not stupid 14:23 MTDiscord I don't see dismantling the dev wiki as a prerequisite to the minetest docs project at all. The point of the endeavor is to create a documentation source that promotes consistency, detail, and usability, which is far more than the current documentation can say. Progress is slow but we are volunteers. At least someone is trying. 14:37 Zughy[m] it's not a prerequisite, it's about keeping things clean 14:46 MTDiscord That was never something Minetest cared about before 14:49 sfan5 and? 14:50 MTDiscord Thank you for your useful input 14:52 MTDiscord I really do not appreciate people complaining about a volunteer project being slow and pointing out extraneous nearly-irrelevant issues instead of coming up with strategies or questions to get the project moving along 14:52 sfan5 I was asking you what you meant by that 14:52 MTDiscord That was far from clear, I apologize 14:53 sfan5 is it "we never cared about keeping the dev wiki clean so we shouldn't now"? 14:54 MTDiscord It meant "why are you claiming cleanliness as a reason this project shouldn't exist when that has never been an issue in the past" 14:54 sfan5 are we talking about the dev wiki now or minetest_docs? 14:54 MTDiscord The latter 14:55 sfan5 I thought cleanliness was about why 90% of the dev wiki should be deleted 14:55 MTDiscord That is fine and not what I an referring to 14:56 sfan5 you were referring to the earlier comments then, that wasn't clear 14:57 MTDiscord Zughy understood but I apologize anyway then 14:57 MTDiscord I figured my message referring to both projects and the prior messages would have implied that but perhaps it didnt 14:58 MTDiscord But bow that we are hopefully on the same page 14:58 MTDiscord Now* 14:58 MTDiscord My takeaway from zughy was "the minetest docs project should not have been approved because you can't put consistent time commitment into it and the dev wiki should have been dismantled first", which is simply ridiculous 15:02 MTDiscord My hope is that I gravely misunderstood, but that doesn't appear to be the case yet 15:07 sfan5 that is basically what I understood too 15:28 celeron55 i don't think it matters in which order those things are done 15:29 celeron55 the end result is the same - cleaned up dev wiki + better docs elsewhere 15:30 celeron55 as long as you're working towards that end result it's what we want 15:32 celeron55 some people like to work in a way where they first clean up and then start a new thing 15:32 celeron55 the way i tend to work is i just start a new project on the remains of others 17:17 Krock rubenwardy: how do I test #12703 ? 17:17 ShadowBot https://github.com/minetest/minetest/issues/12703 -- [NoSquash] ContentDB: Fix crash when trying to overwrite a package by rubenwardy 17:18 rubenwardy mkdir mods/basic_materials && touch mods/basic_materials/init.lua 17:18 rubenwardy attempt to install basic_materials, there will be a dialog select overwrite, it will crash on master 17:19 rubenwardy will add to PR op 17:22 Zughy[m] GreenXenith: my point was "if you start something, you're now responsible of it". If 4-5 people started minetest_docs and it's been 6 months nobody touched anything but one person (who's already pretty busy with other stuff), that's poor thinking. Because we've ended up with yet another unfinished docs and, if I hadn't mentioned the topic earlier, it'd probably have remained buried for some more time 17:24 Zughy[m] and now there's probably someone that wants to write to work on it because of me mentioning. Imagine if CDB or MTirrlicht worked following the same principle since it's volunteering 17:24 Zughy[m] *to want to work / mentioning it 17:26 rubenwardy Good thing that ContentDB is finished software and has zero bugs 17:35 nrz rubenwardy, unless next dependency release for bugfix/secfix ? 17:57 Zughy[m] celeron55: are you sure it's better to edit these pages (https://dev.minetest.net/Category:Proposed_for_deletion) with external links rather than completely deleting them? 17:58 Zughy[m] Deleting require less time ? 17:58 Zughy[m] *requires 17:58 Zughy[m] also https://dev.minetest.net/MinetestWiki:TODO#Delete_pages_marked_for_deletion 18:04 Krock rubenwardy: game overwriting gets stuck while downloading 18:05 Krock wait sorry, my bad. it works 18:15 celeron55 Zughy[m]: yeah that's some work. maybe they should be just deleted 18:16 celeron55 Zughy[m]: you are zughy on the devwiki, right? 18:16 celeron55 Zughy* 18:17 Zughy[m] yep 18:18 Zughy[m] I can make them go away before tomorrow 18:19 celeron55 i gave you the admin privilege for 1 month and a permanent moderator privilege 18:21 celeron55 i actually have no idea what the moderator privilege allows 18:22 celeron55 so, i don't have time to go through all of those pages marked for deletion 18:22 celeron55 but IMO any pages that clearly are inferior documentation than what's available officially elsewhere (including the modding book and whatnot) can be deleted 18:24 celeron55 the next task after that is to figure out how to lead those seeking the info into the proper places 18:27 rubenwardy what's the difference between `str_split` and `split` in util/string.h? 18:30 Krock rubenwardy: split allows escaping delimiters 18:30 Krock just by looking at the code 18:31 celeron55 it seems to allow escaping anything 18:31 celeron55 including the delimiter 18:31 Krock also sorry for yesterday. I should've continued the meeting, but was probably too tired at that point. If we could do the meeting earlier next time, that would be great. 18:32 rubenwardy is there a unit test util for comparing arrays? 18:33 Krock memcmp using the first array index then calculate the length? 18:33 Krock ah.. for strings that won't work tho 18:33 rubenwardy it's a std::vector 18:33 celeron55 i think split() should be renamed escape_split() or something 18:34 Krock rubenwardy: find_if, then use a lambda to compare indices? 18:34 rubenwardy minetest could do a lot better at helpful utils for code reuse 18:35 celeron55 construct std::arrays with the std::vector elements and then just ==? 8) 18:36 celeron55 C++ is silly, there are so many wrong ways of doing things 18:36 sfan5 you can use std::equal between different types of containers 18:36 sfan5 I think 18:36 celeron55 yeah i was going to suggest that as the serious suggestion 18:37 rubenwardy I actually only need one test case with 3 elements, so I just did it manually. Only 4 lines 18:37 celeron55 std::equal and a size comparison 18:37 celeron55 or does it compare the size also? 18:38 celeron55 i don't think it does 18:39 Krock size comparison would be another simple assertion anyway 18:39 celeron55 it's easier to justify making collections of essentially random utility functions in a solo project than a team project 18:40 celeron55 and easier to do it to begin with as you remember what there already is and isn't 18:40 celeron55 not that you actually need any for that 18:40 rubenwardy I think good util functions is vital when it comes to bigger projects 18:41 rubenwardy context: I'm adding str_split_trim and Settings.getList so that I can read comma separated settings easier 18:41 Krock ^ LGTM 18:42 celeron55 i think the rule would be, add anything you need as long as you also add a unit test for it and if you need lots of header dependencies try not to add them to a header everything else uses 18:43 Zughy[m] Krock: should I reschedule it at 18:00 UTC+0 rather than 19:00? 18:44 Krock well, it's up to Hugues and x2048 who IIRC would like to have it not much earlier 18:45 Krock HuguesRoss, @x2048 what would you think about moving the meeting time 1 hour earlier, i.e. 18:00 UTC? 18:46 HuguesRoss No opinion 18:46 HuguesRoss 19:00 is ~3PM for me, so as long as long as it's after 17:00 I don't really mind 19:08 MTDiscord I don't mind either 19:09 Zughy[m] I'll edit then 19:25 Zughy[m] celeron55: I think some updates broke the wiki theme, see all the `` syntax all around the place https://dev.minetest.net/minetest.get_craft_result 19:25 Zughy[m] *all over 19:26 celeron55 yeah i managed to set up the same extensions but for some reason they just don't work the same anymore 19:27 celeron55 that's what 10 years of updates does to mediawiki 19:33 rubenwardy merging #12703 in 10 19:33 ShadowBot https://github.com/minetest/minetest/issues/12703 -- [NoSquash] ContentDB: Fix crash when trying to overwrite a package by rubenwardy 19:36 rubenwardy and #12693 19:36 ShadowBot https://github.com/minetest/minetest/issues/12693 -- Apply DPI Scaling to GUIModalMenu by ElliottLester 20:05 Krock rubenwardy: 10 what? hours? 20:13 celeron55 years obviously 20:20 rubenwardy haha typically. I realised I hadn't ate so went to make food 20:33 Zughy[m] these are all the methods imho worth keeping: https://dev.minetest.net/Category:Methods appguru if it might help you for the other docs 20:39 rubenwardy the dev wiki language selector has broken 20:44 Zughy[m] I think this page is kind of useless: https://dev.minetest.net/How_to_communicate 20:48 rubenwardy I think tht's IRC legacy 20:48 rubenwardy but we have https://wiki.minetest.net/IRC#Rules 20:49 rubenwardy I agree with deleting that 20:50 Zughy[m] done