Time Nick Message 17:20 sfan5 std::hash>{}(it.first) 17:20 sfan5 somehow it feels stuff shouldn't be so complicated 17:23 Krock depends on what you'd like to achieve. A naive approach might be more readable. 17:37 sfan5 it's just debug code but it's a bit surprising you can't just invoke std::hash without specifying the template type 17:39 Krock it's probably generic enough to get confused between notations if it's not nailed down 17:44 Krock Meeting in 15 minutes, apparently. 17:45 [MatrxMT] o/ 17:46 MTDiscord Krock: I would have guessed that the problem is that it isn't specialized for const types 17:56 MTDiscord Though I think C++ doesn't really have good template argument deduction for anything that isn't a function at all? 18:00 Krock Meeting start. Pinging those who reacted. v-rob @v-rob Zughy @luatic 18:00 MTDiscord also @cx384 18:00 Krock oops. yes. 18:01 Krock > #15322 18:01 ShadowBot https://github.com/minetest/minetest/issues/15322 -- Minetest → Luanti rename checklist 18:01 Krock > Finish renaming Minetest -> Luanti 18:01 Krock There's not much that we could do there. That's mostly administrative work 18:02 Krock i.e. for c55 or people who've got more access than regular developers 18:04 MTDiscord On that list, I think there are roughly 3 things for core devs to think about: 18:05 MTDiscord 1. Do we want to rename/migrate/move the GitHub repo/org? (low priority but we should consider it) 18:05 MTDiscord 2. What do we do about the PPA? Add a new luantidevs one? Publish a new luanti package on the minetestdevs one? 18:06 MTDiscord 3. We also should do the grunt work of resubmitting on Flathub and doing an EoL rebase eventually 18:06 Krock for 1. it would make sense for me but I have no urgent need to have that changed. 2. is best solved with a transitional package (still "minetest" but with luanti as a dependency) 18:07 MTDiscord I believe (2) is solved the other way around: Have a new luanti package which "provides" minetest. 18:07 Krock hmm 18:08 [MatrxMT] Sorry, here I am 18:09 [MatrxMT] About that point I mailed c55 ~10 days ago kindly asking if he could tackle the points we have no control about, but I received no answer.. 18:10 celeron55 wait what 18:10 MTDiscord bonjour 18:10 [MatrxMT] November 25 celeron55 18:10 MTDiscord something went to someone's spam folder i'm guessing ;) 18:16 celeron55 oh yeah it was just a reference to the discussion thread 18:17 celeron55 ...i mean the issue 18:22 Krock If someone's willing to take care of the Flatpak or PPA topics, that would be great. It's not like we're in a hurry. 18:22 Krock Are there any issues or PRs that you'd like to discuss? (in general) 18:25 MTDiscord yes 18:27 MTDiscord as for the flatpak, i might do it (i already looked into it a bit), but i can't guarantee i'll have the time. i'm too unsure to do anything about the PPA tbh. 18:28 MTDiscord i'm wondering what we're doing about physics 18:29 Krock I had in mind to re-test that Soon (TM) 18:29 MTDiscord ah, just saw kno10 dropped the acceleration-based formulae as well. good, i was worried for a bit there because i had hit a bit of a dead end. 18:29 Krock It's not so good because it'll be imperfect by definition. But at this point I suppose it's good enough to go in small steps 18:30 Krock (i.e. iterative steps towards a hopefully better working collision engine) 18:30 MTDiscord Krock: I know, but the acceleration-based formulae are a mess. If you try to be accurate, a lot can happen. It probably also hurts performance significantly. 18:33 Krock sqrt is not efficient. I know that. Perhaps we can somehow find a solution to calculate the dtime steps more accurately 18:33 Krock (beforehand, rather than for each colliding face) 18:33 MTDiscord Yeah I think ultimately running the physics at a higher rate, but still with a linear interpolation, will be the way to go 18:36 Krock if it works reasonably precise, then we might be able to increase the step size. I believe that's currently set to very low to catch such edge-cases. 18:36 MTDiscord celeron55: "This obviously means someone has to configure the github pages for each accordingly" - it seems possible that that someone would have to be you again though I'm not sure 😬 18:36 MTDiscord I do not have access to the repository settings as far as I know 18:36 celeron55 for github pages, it's the file called CNAME 18:37 celeron55 i don't think that has any special authorization. it's just a file in the repo 18:37 MTDiscord aha, thanks 18:38 MTDiscord okay so I'll just edit that to blog.luanti.org and see what happens 18:38 celeron55 i imagine that will work for the blog 18:38 celeron55 as for api.minetest.net... i have no idea how that thing works 18:39 MTDiscord yay seems to have worked! https://blog.luanti.org/ 18:39 rubenwardy https://blog.minetest.net/ is now broken 18:39 MTDiscord however https://blog.minetest.net/ does not work anymore 18:40 MTDiscord i thought that ought to redirect to blog.luanti.org hrm 18:40 celeron55 yeah i need to do that now 18:41 celeron55 good thing that i didn't really prepare to do it 18:41 celeron55 8) 18:43 MTDiscord I think api.minetest.net is coming from GitLab pages actually 18:43 celeron55 just a moment 18:44 MTDiscord or wait maybe that was fake news 18:44 celeron55 uh... the dns record for api.minetest.net is CNAME minetest.github.io, though 18:45 MTDiscord "The CNAME file can contain only one domain. To point multiple domains to your site, you must set up a redirect through your DNS provider." 18:45 rubenwardy It's GitHub pages as well, built using an action. I believe you change the domain in the GitHub pages settings 18:45 MTDiscord yeah that's one way to do it, but we can just edit the CNAME file in the minetest repo as well 18:46 MTDiscord as i just learned ;) 18:46 rubenwardy If there is one 18:46 rubenwardy Which there is, so do that 18:46 celeron55 https://blog.minetest.net/ will now redirect to https://blog.luanti.org/ as soon as the viewer's DNS updates 18:46 rubenwardy I thought the confusion was over there not being a cname file 18:46 MTDiscord (I was misled into thinking that it was GitLab pages because if you go on GitLab and then follow the link labeled "GitLab pages", it brings you to api.minetest.net) 18:47 MTDiscord I think api.minetest.net was GitLab in the past? 18:47 MTDiscord okay so i'll just edit the CNAME file in the minetest repo to api.luanti.org as well 18:48 celeron55 "I believe you change the domain in the GitHub pages settings" <- where is "GitHub pages settings"?" 18:48 celeron55 -" 18:48 celeron55 ah so it's in the main repo 18:49 ROllerozxa the CNAME file turned out to actually not be used when github pages deploys using github actions 18:49 MTDiscord huh where does it get the domain from then 18:49 celeron55 well i'd love to change it then, if i knew where it is 18:49 celeron55 or someone else 18:50 celeron55 i'll set up the redirect for this now 18:50 ROllerozxa minetest/minetest -> settings -> pages -> custom domain 18:50 ROllerozxa it's quite weird honestly, normally when it's just deploying without actions it will sync CNAME with the repo 18:50 MTDiscord huh i thought that was just another way of editing that file 18:50 ROllerozxa yeah that's how it behaves if you don't deploy with actions 18:51 ROllerozxa (i.e. with legacy jekyll or just pushing static files) 18:51 ROllerozxa so the CNAME file in minetest/minetest could be deleted altogether 18:52 celeron55 i now set that custom domain setting 18:52 ROllerozxa awesome 18:53 ROllerozxa yeah now api.minetest.net redirects to api.luanti.org and it works for me 18:56 celeron55 i'm setting up these redirects to work so that the dns points to the server that runs the forum and wikis and its main nginx is configured with ssl, acme and a 301 redirect for each 18:56 celeron55 seems like the only reasonable way to do it 18:58 celeron55 well, if i was using cloudflare or something, then this would be done there, but as long as this is the closest thing to the internet that i have, then this is the way i have to do it 19:05 [MatrxMT] It seems like dev.luanti.org does not redirect yet? 19:05 MTDiscord hmm http://api.luanti.org/ gives me "there isn't a github pages site here" 19:08 MTDiscord it exists for me 19:08 MTDiscord probably caching 19:08 MTDiscord try doing ctrl+f5 19:09 MTDiscord (github pages can cache things quite aggressively) 19:09 MTDiscord alternatively alt f4 for maxium enjoyment 19:09 MTDiscord i did try all of that 19:09 MTDiscord no dice so far 19:10 MTDiscord assuming you have another browser on your machine, try it 19:10 MTDiscord same issue on chrome 19:11 MTDiscord well whatever, since it works for you i'll try again tomorrow and chalk it up to DNS weirdness if it works then 19:11 Krock will merge #15415, #15495 and #15486 in 15 minutes. My new main menu PR can surely wait a little so that people get the chance to comment on it. 19:11 ShadowBot https://github.com/minetest/minetest/issues/15415 -- Add `core.spawn_tree_on_vmanip` by cx384 19:11 ShadowBot https://github.com/minetest/minetest/issues/15495 -- [NO SQUASH] Mapgen: Fix Y biome boundary regression by SmallJoker 19:11 ShadowBot https://github.com/minetest/minetest/issues/15486 -- Adopt #13446 server favorite button by siliconsniffer 19:14 MTDiscord Haven't tested it extensively but does as promised. 19:17 sfan5 testing welcome for #15531 19:17 ShadowBot https://github.com/minetest/minetest/issues/15531 -- [no squash] Group sparse mesh buffers over entire scene for rendering by sfan5 19:18 Krock how to test: you could upload a minimal world configuration to compare before/after. Also server spawn points can be interesting. 19:22 Krock ^ also the variables in clientmap.cpp could benefit from some renaming. it's not entirely clear what they're supposed to do 19:24 sfan5 the code is not production-quality (yet) 19:25 Krock merging the PRs.... 19:28 sfan5 https://gdcvault.com/play/1020791/ this task is supposedly good 19:28 Krock done 19:28 sfan5 talk* 19:29 celeron55 @y5nw yes, i didn't touch devwiki yet. it's not up to date anyway. very low priority 19:29 celeron55 everyone is trying to drop it as fast as they can 19:30 MTDiscord checked off api.minetest.net since thats done now 19:36 Desour sfan5: std::pair has a second_type member type, fyi, no need for the decltype :) 19:45 [MatrxMT] RE "everyone is trying to drop it as fast as they can": what do you mean celeron55? 19:50 celeron55 mostly just that it's not a recommended source anymore, and it's not being edited anymore 19:51 [MatrxMT] can we get rid of it, saving just a couple pages? 👀 19:51 celeron55 www.luanti.org does link to the "modding intro" page still though 19:51 [MatrxMT] forum.luanti.org dead 19:52 MTDiscord took forever as usual, but loads for me 19:52 celeron55 which means, the official process of introducing people to the modding api probably needs to be looked at 19:53 celeron55 there obviously needs to be a portal where people can easily begin, and... do we not have anything better than this yet? https://dev.minetest.net/Modding_Intro 19:54 MTDiscord just take the content from that, and make it a subpage of luanti.org perhaps? 19:56 MTDiscord or move it over to main wiki 19:56 [MatrxMT] celeron55: food for thoughts about the two wikis: https://forum.luanti.org/viewtopic.php?t=31080 19:57 Krock Why do you want to change everything? 19:57 [MatrxMT] My role is to keep things in order 19:57 [MatrxMT] celeron55: I was thinking more about the dev wiki in terms of a place for explaining/documenting the development process, since that's (I recall) the only place I found (IMO useful) information when you asked me about becoming a coredev 19:58 Krock > To avoid seeing all the previous discussions disappear, an "Archive" section can be added 19:58 rubenwardy https://dev.minetest.net/index.php?title=Special:AllPages&hideredirects=1 19:58 Krock I'd consider this as a mess 19:59 rubenwardy a section contains multiple sub forums, he's not suggesting merging into a singhle forum 20:00 rubenwardy example being "General" and "User-contributed content" 20:00 celeron55 @y5nw yes, that's the only real use for it at the moment and it will exist for that until it's no longer necessary for that 20:00 [MatrxMT] (I linked it mostly for the wikis, but if you want to comment everything please do 👀) 20:00 celeron55 @y5nw that purpose could easily be replaced by a git repo with the same content, though 20:01 MTDiscord c55: so if i convert the dev wiki to git+markdown, it would be accepted to transfer it into the org+point domain at it? 20:03 celeron55 yes please. i hate having to drag the devwiki around 20:04 MTDiscord ok, will work on that 20:05 ROllerozxa I guess it should be decided what the devwiki should be too 20:06 ROllerozxa is it only for the C++ engine and associated documentation or for modders too 20:06 MTDiscord fair disclousure, im ignoring anything like https://dev.minetest.net/InvRef that is already marked for deletion 20:06 ROllerozxa yeah any stubs and other such crap should be ignored 20:06 MTDiscord roller: just going to focus one getting stuff converted as a first step 20:06 ROllerozxa sure 20:08 celeron55 only the pages that are providing extra insight over what's on api.luanti.org should be saved 20:10 Desour @wsor: here's a markdowned part: #12749 20:10 ShadowBot https://github.com/minetest/minetest/issues/12749 -- Move code style guidelines into repo by Desour 20:12 celeron55 i agree with Zughy that overall there's a lot of pruning to do and things need to be re-concentrated to a reasonable extent. basically cut off the dead branches and point links to the healthy ones 20:17 MTDiscord okay api.luanti.org works for me now 🎉 20:18 MTDiscord devwiki progress is being worked on here: https://github.com/wsor4035/devwiki