Time Nick Message 00:03 v-rob libunistring is an optional dep of gettext 00:07 v-rob OK, here's what I'm going to do: I'm going to use ICU for my development for now, since it's entirely available everywhere, and I'll attempt to write things in such a way that I can change to the GNU libraries if need be. 00:08 v-rob I guess it's a good thing c55 took a quick look over everything. I didn't think of those things... 00:52 MTDiscord Does anyone know of an issue offhand with requiring spatial index on all our currently supported platforms? 01:30 Calinou ICU + harfbuzz will give you best results, but harfbuzz isn't small 01:31 Calinou ICU data also takes a few megabytes, but that should be fine for Minetest 03:26 kilbith hey v-rob, I'm still waiting for the formspec_version bump; that's important for me 03:27 v-rob Thanks for reminding me; will do soon 10:09 celeron55 v-rob: i don't think the situation with libunistring is so bad you have to avoid it, but you do need to plan what the recommended windows build process will be 10:09 celeron55 (and other platforms also) 10:09 celeron55 if you don't feel like figuring out build processes then icu probably makes more sense 10:18 celeron55 i'm guessing building libunistring for windows isn't much of a hassle 10:18 celeron55 and for android? no idea 10:59 sfan5 x2048: sure you can add that to the milestone, but that's mainly just we don't forget 11:00 sfan5 since it has my and your (implicit) approval you can also just merge it right now 11:01 sfan5 huh I thought the "announce merges 5..15 min before doing them" was actually written down 11:01 sfan5 but the written down part only applies to trivial patches pushed spontanously 11:02 sfan5 (expected it to be here https://dev.minetest.net/Git_Guidelines) 11:06 celeron55 i think that's intentional 11:07 sfan5 I never knew! 11:07 celeron55 it of course doesn't hurt to announce a merge 11:08 celeron55 especially if it's something someone might want to discuss still 12:10 x2048 sfan5: thank you, then I'll merge #11961 now. 12:10 ShadowBot https://github.com/minetest/minetest/issues/11961 -- Reset override material in anaglyph by x2048 15:25 sfan5 pushing https://github.com/minetest/minetest/commit/b02b381af26c792d7f7feaa93198bb7efb2512de in a few minutes 17:45 Pexin ls 17:52 erlehmann wrong window Pexin 17:53 Pexin OH NO! 17:53 Pexin :] 19:42 v-rob celeron55: I'm approximately the worst at working with build systems. For the first two years of my contributing, my workflow was 1) push work to GitHub, 2) pull onto a Linux VM, 3) recompile from scratch using a one-liner script, all because I couldn't figure out how to compile on Windows. I've improved since then, but I'm still pretty iffy in general. So unless someone else finds a way to build libunistring, I'm sticking with ICU :) 19:42 celeron55 i think sfan5 could probably come up with some useful opinions about this but i don't know if he wants to focus on this right now 19:43 v-rob Anyhow, it's not a huge deal either way. I can work with ICU well enough, and it's not a bad choice in any case. 19:43 Krock isn't there vcpkg support now to ease library integration on Windows? 19:44 v-rob libunistring isn't on it 19:44 v-rob Well, it is, but not for Windows 19:44 Krock where do you need that dependency? 19:44 sfan5 figuring out how to build libunistring when it's needed won't be a problem 19:45 v-rob It was one of the options for the Unicode library for the GUI replacement 19:45 sfan5 it not being on vcpkg is bad for MSVC users because it invalidates the point of those build instructions (easy to set up) 19:45 v-rob Yeah 19:46 sfan5 https://github.com/archlinux/svntogit-packages/blob/packages/libunistring/trunk/PKGBUILD#L31-L33 see this looks easy enough ;) 19:46 v-rob If we change libraries, most of the Unicode library function calls are restricted to a single class I'm writing right now that does everything Unicode, so it wouldn't be too difficult to change. 19:48 v-rob See, it's easy to build on Linux. But I'm not on Linux, unless I want to go back to working on a Linux VM :) 19:48 v-rob I only have a single computer right now, and it's kind of required that I runs Windows because of software I'm required to use. 19:52 v-rob One of these days, I'm going to find all the people who work exclusively on Linux, lock them in a room, and force them to build all their software on Windows without the aid of a Linux command line, and have them see what they're doing to people. 19:54 erlehmann look if you have git, you have sh 19:54 erlehmann and i write most of my software in pure sh 19:55 erlehmann so i will of course win this competition and get the prize! (a windows gaming pc that is bootlocked or something) 19:56 v-rob That would defeat the purpose of the exercise. Just for that, I'd supervise you personally :) 20:00 MTDiscord Doesn't Bash run on Windows nowadays? 20:01 celeron55 well, _linux_ runs on windows nowadays 20:01 celeron55 frankly we're living in an upside down world 20:01 MTDiscord Without pulling in a VM :-) 20:01 v-rob It really doesn't make it any easier in my experience 20:01 v-rob Or maybe the world conspires against me. It's hard to tell 20:02 MTDiscord No, but I'm not learning powershell thank you very much 20:02 MTDiscord /s 20:02 MTDiscord (Sort of /s, I'd rather not) 20:03 sfan5 the choice of MSVC only makes it harder, even if you have a build system that abstracts it away 20:03 sfan5 there is software that downright does not support msvc because of it's lousy C standard support 20:04 v-rob I'd prefer to use MinGW for Minetest, but that didn't go well in the past 20:05 MTDiscord I can definitely build software on a Windows machine, it just won't have Windows on it at the end of the process. 20:06 celeron55 i mean, just use it as a table to set up your linux machine on top of 20:06 MTDiscord I agree that building software intended for Linux on Windows does kinda suck, but based on my experiences back when I ran windows, building software intended for Windows on Windows kinda sucked. 20:07 v-rob :D 20:33 sfan5 I know nobody really cares about MTG but https://github.com/minetest/minetest_game/milestone/2 needs reviews *or* agreement to revoke the "two approvals" rule 20:36 celeron55 i have a note about possibly adding appgurueu to the mtg team, which i guess could solve that. is it something you'd like? 20:39 sfan5 that's an okay solution 20:39 sfan5 just potentially a bit weird when MTG is on its way to the grave 20:40 celeron55 well i mean, grave diggers 20:41 freshreplicant[m Is it really though? On the serverlist too? 20:41 sfan5 development-wise 20:41 sfan5 people won't stop using it anytime soon 20:43 MTDiscord i mean the more popular servers run custom mtg or override a lot of it 20:44 celeron55 it will probably always be a modding base for that type of game 20:44 celeron55 in a giant network of forks 20:45 freshreplicant[m If it's possible to make performance improvements or optimise it, I don't think anyone would say no to that. 20:46 freshreplicant[m This is MTG: https://www.youtube.com/watch?v=jYcPBE5PXhs 20:48 celeron55 mtg already died and was resurrected once, a long time ago. at that time the development transferred from the core team to the (new) game team 20:49 celeron55 i'm skeptical of any third life 20:49 celeron55 altough all the forks basically are where it lives 20:49 freshreplicant[m It's not dead though. As much as some people want it to be. 20:49 celeron55 show me a proposal and a team, and i'll look at it 20:49 freshreplicant[m It's still a huge chunk of the forums, Youtube (counting modded MTG). 20:50 v-rob If MTG is resurrected again, we'll probably get a free rant from Wuzzy :) 20:51 freshreplicant[m Why not just get more people to approve the improvements/fixes and minor tweaks people already want to make? 20:51 freshreplicant[m Everything will get a rant from people, I'd almost be worried if it didn't. 20:52 celeron55 can you get those people? 20:52 sfan5 improvements, fixes or tweaks are not the problem (even if we had the people) 20:52 celeron55 i think not that many would be interested 20:52 sfan5 MTG is missing a direction as a whole 20:53 MTDiscord ^this 20:53 freshreplicant[m I don't necessarily mean making it into some massive game. I'm just talking about saying 1) PRs welcome 2) And like sfan5 said, make it easier to approve PRs. 20:53 MTDiscord you cant please anyone 20:54 MTDiscord . s/anyone/everyone 20:54 MTDiscord so no one is happy 20:54 MTDiscord its not a game its a modding base. Better to design a game, and steal code from it 20:54 v-rob It has been missing a direction for about 12 years now, which is the length of Minetest's existence 20:54 MTDiscord textrues too, if youre lazy 20:54 celeron55 arguably it had a direction for about half a year 20:55 MTDiscord besides, it would be much better fixing mte pr merging situation given sfan5 is basically defacto the main person and seems he struggles to get second approvals now 20:55 sfan5 "now" 20:55 MTDiscord for a while 20:55 freshreplicant[m How many PRs have been made recently, and issues, despite the pinned "This is dead, move along" issue? 20:55 MTDiscord go look yourself lol 20:56 sfan5 does anyone agree that #11922 can go into 5.5 despite being a feature? 20:56 ShadowBot https://github.com/minetest/minetest/issues/11922 -- Allow resetting celestial vault elements by leaving its arguments empty by Zughy 20:56 freshreplicant[m It was a rhetorical question. 20:56 celeron55 well, i'll wait until i hear from appguru 20:56 celeron55 the 5.5.0 PRs need to get handled somehow 20:57 celeron55 if more than that is to happen, we'll see 20:57 MTDiscord issues: jul, oct, 15 days ago prs: jul, nov, 3 days ago seems not much freshreplicant 20:58 celeron55 the mtg team is autonomous and fully trusted to make literally any decision about mtg so as long as there are active people there you don't really need to convince me about anything 20:58 freshreplicant[m Still happening while it's dead, with a slim change of approval. 20:59 celeron55 but there are reasons for the current situation 20:59 celeron55 and the main reason was mentioned by sfan5 already 21:01 freshreplicant[m Yeah, I said what I said in support of making it easier for PRs to get approved. 21:01 rubenwardy v-rob: I moved to linux precisely because compiling on Windows was/is a pain 21:01 freshreplicant[m I'm not a huge MTG advocate anymore, because I know how determined people are to double tap it and bury it six feet under, but I do sometimes think it's funny how everyone pretends it's dead already. 21:02 freshreplicant[m When it still amounts to a ton of activity in the MTE universe. 21:03 appguru celeron55: Here I am 21:04 appguru I would like to be on the MTG team 21:04 appguru MTG should pick a direction other than "maintenance-only" sooner or later 21:05 appguru But IMO, first of all, a massive cleanup would be in order 21:05 sfan5 narrator: it won't 21:05 appguru lol 21:05 MTDiscord just let it die 21:05 MTDiscord its screwed from the name, being default in the engine, code design, etc 21:06 appguru Jonathon: Unfortunately we have no satisfying replacement yet 21:06 appguru Especially considering all the MTG-tailored mods & all the MTG forks 21:06 MTDiscord it doesnt need to be replaced tbh 21:06 sfan5 well the 'being default' situation was supposed to be resolved (by the engine!) too but progress on that is stalled 21:06 appguru I'm fine with MTG ceasing to be the default, that's a good idea 21:06 MTDiscord Once you pick a direction and start moving in that direction, you'll basically need to leave the MTG name behind, which is probably just as well. 21:06 appguru But we can't kill it off 21:07 MTDiscord your also stuck in compatibility hell 21:07 sfan5 @Jonathan to many people Minetest is still a means to have a vaguely minecraft-inspired/like game, that wish won't just disappear 21:07 appguru Theoretically MCL is supposed to fill that spot 21:07 sfan5 turns out people don't want exactly Minecraft either ;) 21:07 appguru Well, MCL is quite popular 21:07 MTDiscord sfan5: i was refering more to people confusing minetest game with minetest engine, which is very common 21:07 appguru It's also quite messy though 21:07 sfan5 true 21:08 appguru There's a forest of forks 21:08 appguru Many are incomplete and/or buggy or unmaintained 21:09 appguru If I'm not mistaken, most servers on the list run heavily modded MTG 21:10 MTDiscord if you really want to continue it, honestly should build a compatibily.lua file into each mod that translates each old api call into a new one, however that will still break mods depending on texture names, node names, etc 21:10 appguru True 21:11 appguru As of now, 126 servers say that they use mod named "default" 21:11 appguru So it's pretty safe to assume that there are at least 100 MTG servers 21:11 wsor and i bet you over half of those are modified default 21:11 appguru No doubt 21:12 MTDiscord Modified MTG is still MTG, from a certain perspective. Changes made to the official upstream MTG will either need to work their way down into private forks, or those forks will start running into compat issues with other mods they do want to update. 21:12 freshreplicant[m Also, Minetest finally got on Wikipedia. 21:12 MTDiscord MTG is "dead" as a code project but still pretty dominant as a "standard/protocol" 21:12 wsor very old news freshreplicant[m 21:13 freshreplicant[m First line: "Minetest is a free and open-source sandbox video game and game creation system developed by a team of volunteers, with significant contributions from the community.[7]" 21:13 rubenwardy got on wikipedia, but incorrectly 21:13 freshreplicant[m I know, my point is, it's still thought of a game by many people. 21:13 wsor if you want to continue mtg, your probably best off making a fork and breaking tons of compat in it 21:14 MTDiscord Yes, a lot of people think of it as a game, but that's just going to take longer to fix. 21:19 freshreplicant[m You could fork it, break compat, give it a total makeover. Or you could keep it in its current role, but make it more streamlined to make and approve those small fixes/optimisations/features that don't break compat. 21:20 MTDiscord if you give it a total makeover id argue that effort is better spent on a new game from scratch anyways 21:20 MTDiscord pull small sections from mtg to speed your dev up 21:22 appguru But MTG is already established 21:22 appguru So maintenance is the bare minimum 21:29 freshreplicant[m I agree. 23:59 MTDiscord Why not take the direction you think MTG should go, and go make something with less cruft from the ground up? MTG is most commonly a modding base, so making a new one thats actually decent would likely gain some traction 23:59 MTDiscord But MTG cannot be everything it wants to be, or needs to be