Time Nick Message 11:56 * mtvisitor will review some document of "farming redo" mod later (if possible). 11:57 mtvisitor #14865 11:57 ShadowBot https://github.com/minetest/minetest/issues/14865 -- Remove mod_translation_updater.py and move it to minetest/modtools by Zughy 12:01 mtvisitor do you know what is a good practice to change the username locally in singleplayer mode or multiplayer mode for minetest game? 12:05 Krock mtvisitor: singleplayer is hardcoded and cannot be changed 12:05 Krock and names cannot be changed in general due to meta field references 12:06 mtvisitor ok, thank you very much for your answer and help. 12:06 Krock chest "owner" fields do "hardcode" your username into it, thus it's impossible to rename them all 12:07 Krock (same logic applies to almost all player-specific protections or other configuration with usernames) 12:07 mtvisitor Krock: sorry for 1 more question. 12:08 Krock mtvisitor: do not be sorry. do ask. 12:08 mtvisitor do you know how to disable pvp on multiplayer mode locally ? 12:08 Krock main menu settings (checkbox) to save the setting into the world's configuration file OR 12:09 Krock you edit "world.mt" with "enable_damage = false" 12:09 Krock hmm.. looking what's about "enable_pvp" 12:10 Krock "enable_pvp" is a global setting and cannot be changed on a per-world basis 12:10 mtvisitor ok, i will check these configuration(s) later (SCM). thank you very much for your help. 12:11 Krock PlayerSAO::punch for reference 12:12 mtvisitor i will get offline.thanks. 12:12 mtvisitor good night. 12:14 Krock good afternoon 18:23 MTDiscord Good afternoon too 21:11 sfan5 PSA: https://content.minetest.net/packages/sfan5/minetest_classic/ 21:15 ireallyhateirc minetest classic(TM) 21:15 ireallyhateirc are rats included? I love rats 21:15 ireallyhateirc oooh there's a rat in the screenshot 21:16 sfan5 it wouldn't be a real port without rats, oerkkis, dungeon masters and the fourth mob I forgot 21:19 celeron55 love it. "my painstaking recreation of... well i don't remember anymore" 21:23 celeron55 comparing that to the original in terms of the engine would be brutal. not giving video ideas, but that's definitely a video idea 21:24 ireallyhateirc btw I have a question. why did you add the xyz -32 grid offset instead of just spawning sam at xyz +32 ? 21:25 ireallyhateirc would be much easier if mapchunks started at xyz 0 21:27 celeron55 i don't remember. i agree it seems weird 21:28 celeron55 i think the idea probably was that chunks wouldn't really matter and they might as well not exist as their size and position could be anything 21:28 celeron55 to the extent that moving the spawn would matter more 21:29 ireallyhateirc doing the grid trick makes Sam spawn in the middle of a mapchunk, but otherwise it's pretty useless 21:29 ireallyhateirc and gave me funny moments when I started my mapgen adventure 21:30 celeron55 it was important to make the spawn happen in the middle of a chunk, as back then generating a chunk took considerable amount of time. it made for a terrible spawning experience if 4 had to be generated 21:30 ireallyhateirc though can't blame you, I also wrote weird code I regret now in my Minetest mods :D 21:30 ireallyhateirc makes sense 21:30 celeron55 Minetest is old. incredibly old :D 21:32 celeron55 hardware has gotten tens of times faster since then. maybe even 100 times 21:32 celeron55 a common smartphone today is way faster than the laptop i originally developed MT on 21:33 ireallyhateirc and yet Windows 11 still loads 15 minutes on a HDD, some things don't change 21:38 celeron55 people don't realize it, but the reason why Minecraft came about in 2009 was because computers literally weren't capable of running something like that before. in that day's standards, a big dynamic block world like such requires a lot of memory and a lot of processing, much more than is obvious. today, it's trivial to make, but you still can't leave something like it unoptimized, because people 21:38 celeron55 demand so much more from it - more rendering range, shorter loading times, more content. many of the optimizations from back then keep MT relatively competitive 21:38 ireallyhateirc yeah voxels are hard for what they deliver 21:38 ireallyhateirc GPUs love triangles and hate cubes 21:39 celeron55 yes and while you have lots of RAM today, processing through a lot of it is still slow 21:39 sfan5 in an alternate reality GPUs became great at rendering quads and we moved away from triangles 21:40 ireallyhateirc there's an iteresting wip game where the guy made voxels tiny, though don't know technical details 21:41 ireallyhateirc https://www.youtube.com/@johnlin9665 21:41 ireallyhateirc something something OpenCl? don't remember 21:41 ireallyhateirc probably proprietary though 21:42 celeron55 well, i'm going to be the guy who asks: is that a good game? probably not. a nice tech demo though 21:43 ireallyhateirc not saying it's good, but small voxels are impressive 21:45 celeron55 the vibe you always get, though, is "i wonder how loud the GPU fans are while doing that" 21:46 ireallyhateirc yeah also until it's released, it could be simply a Blender render lol 21:47 ireallyhateirc also there was no activity on youtube or twitter since 2020 21:52 ireallyhateirc found something from 2021, the guy himself on his blog talking about voxels: https://voxely.net/blog/the-perfect-voxel-engine/ 22:25 ireallyhateirc found a talk that explains the optimisations, but my 3D programming knowledge is nil so idk if it's right or how much of that is already in Minetest https://www.youtube.com/watch?v=4xs66m1Of4A 22:39 MTDiscord btw sfan5 isn't it a bit risky in terms of setting name collisions to not namespace the "modernize" setting as "mtc_modernize" or similar? 22:40 MTDiscord ireallyhateirc: yeah, minetest already does that, to an extent. you can see how it looks in wireframe mode. fun fact, i made the same suggestion some years ago, and was then pointed (by sfan i think) to the wireframe mode which demonstrated that it was already implemented. 22:42 ireallyhateirc did you fully watch the video? it doesn't stop ad greedy meshing 22:42 ireallyhateirc at* 22:42 MTDiscord ireallyhateirc: no, i only read the title 22:51 ireallyhateirc it does a thing called "binary greedy meshing" and it seems there are some powerful tricks employed there 22:51 ireallyhateirc here the weird stuff starts: https://youtu.be/4xs66m1Of4A?feature=shared&t=754 22:53 ireallyhateirc and after binary greedy meshing he explains even more weird stuff: https://youtu.be/4xs66m1Of4A?feature=shared&t=1276 23:04 MTDiscord ireallyhateirc: alright, i have watched the entire video now 23:04 ireallyhateirc promising or full of holes? 23:05 MTDiscord it is certainly interesting. though i see two issues with applying this to minetest. 23:05 MTDiscord first, his voxel engine is "pure", i.e. he doesn't have the entire drawtype mess we have. 23:06 MTDiscord second, he seems to be pushing a lot of work to the gpu. he's minimizing the amount of faces with his global lattice for example, but maximizing the amount of fragments. i would be worried about how this performs on lower-end gpus - he mentioned a pretty beef gpu. 23:07 MTDiscord and matters like translucent nodes further complicate things. 23:08 ireallyhateirc well, hope that this was at least an inspiration rather than 30 minutes of wasted lifetime :D 23:09 MTDiscord ireallyhateirc: it certainly was inspirational :D 23:11 MTDiscord though i think this is the "work on a cool pure project" kind of inspiration (i kind of feel like writing my own voxel engine now) and not the minetest kind of inspiration. pure problems are a lot more beautiful... 23:19 ireallyhateirc gotta find a "filthy voxels" playlist next time...