Time Nick Message 03:25 MTDiscord Problem? 03:25 MTDiscord https://cdn.discordapp.com/attachments/749727888659447960/1198830725445718137/image.png?ex=65c0551e&is=65ade01e&hm=ff4d837ce2681ea1e072fe7605f6611e8183c517ea414edd9b15be7973e5e6b2& 03:27 MTDiscord Why yes, that is the original trollface I grabbed from the first commit of minetest ever made 03:27 MTDiscord He's colorized in the shader 10:17 MTDiscord Okay so I did some more testing, and now it has a temporary color shader, on top of the colors that are built into the model so I decided to cycle through the channels and then you get this https://youtu.be/qA79q3Nheic?feature=shared 10:20 MTDiscord Now you can have problem? in 3 different colors, fancy 10:46 greeter huh, the minetest server works better if you reboot it more than once every three months. go figure 14:11 MTDiscord are you working on the frontend of mtr? 17:11 whosit hello. It's not clear from documentation what is the lifetime of the particle spawner with time=0. Does it disappear when mapblock is unloaded? If it does, how do I use the `id` that add_particlespawner returns? 17:12 whosit If I just have a table storing those ids, it will fill up with invalid ids? since I can't clear them up on mapblock unloading (there's no callback for that?) 17:23 whosit hmm... looks like I should store the particle spawner id in node meta... 17:24 appguru I don't think particle spawners persist. 17:24 appguru (whereas the point of node meta is that it persists) 17:25 whosit right, and ids will be reused... 17:26 whosit I can't get time=0 particle spawner to "despawn" by leaving the area... 17:26 whosit So, they are kept somewhere? 17:27 whosit the question is: how do I attach a infinite particle spawner to a node? 17:28 appguru whosit: As far as I can see and remember, particle spawners are managed separately from mapblocks. They do "disappear" on clients when they are out of range, but I don't think they disappear on the server (until the server is rebooted, they aren't persisted). If a client gets "in range" again, the particlespawner should appear again. So your table shouldn't fill up with "invalid IDs". 17:29 MinetestBot 02[git] 04cx384 -> 03minetest/minetest: Tool specific pointing and blocking pointable type (#13992) 135958714 https://github.com/minetest/minetest/commit/59587143091a150ac496b20c2cddab9352c6d3fe (152024-01-22T17:27:08Z) 17:30 appguru I think this is somewhat tricky. I would add the "infinite particlespawner" on demand using a LBM (first checking if there isn't a particlespawner already), and I would keep track of these particlespawners in a table [hashed node position] = particlespawner id. I would periodically poll positions in this table - if I get an ignore node, I would delete the particlespawner (mapblock has been unloaded). 17:31 appguru whosit: you can see a rough implementation of this here: https://github.com/spiraling-down/game/blob/master/mods/sd_map/node_particles.lua 17:31 whosit this is partially what I do: https://github.com/mt-mods/abritorch/pull/6/files 17:31 whosit except polling, hmm.. 17:35 whosit thank you appguru, this seems to be what I have to do 17:35 appguru no problem :) 17:37 whosit also, what is the difference between minetest.register_on_dignode and nodedef.on_destruct? seems like on_destruct is more fitting for this? 17:42 appguru IIRC on_dignode is for player-caused actions whereas on_destruct also includes minetest.remove_node or similar. on_destruct is indeed more fitting. 17:42 appguru but it's tricky to keep track of a node without polling if you don't control the entire game 17:43 appguru for example, what if someone replaces your torches using worldedit (voxelmanip)? AFAIK, that won't trigger on_destruct. 17:44 whosit right... but polling will take care of that... 17:45 appguru yep, that's how you do it 17:45 whosit what's the best way to throttle that globalstep I wonder... 17:45 appguru if you want to be clever, you could keep track of the count of torches. the more torches there are, the less frequently you run the globalstep. 17:46 appguru besides that, profile and tweak :) 17:47 whosit feels like the best way would be to do constant amount of work each globalstep, so something like a queue that processes fixed amount of ids... 17:49 whosit why it's so hard X) 17:51 jonadab Then you just have to decide which tasks it's ok to skip when the queue is overfull, versus which ones are worth slowing down the game for in that case. 17:52 jonadab Unless you want to allow the queue to grow arbitrarily large and get further and further behind. 17:54 whosit if only we could just attach stuff to nodes and let MT keep track of it ;) 18:43 whosit another way could be to create an entity that would call on_deactivate and remove the spawners?.. 19:07 MinetestBot 02[git] 04appgurueu -> 03minetest/minetest: Address `set_player_privs` footgun (#14297) 13afc48cf https://github.com/minetest/minetest/commit/afc48cf2242d576aed2371b4763a71fec8739857 (152024-01-22T19:06:03Z) 20:44 chmodsayshello I know I asked this earlier today, but did so while debugging a selfwritten IRC client, so I am not too sure whether the message actually got though :P 20:44 chmodsayshello Anyway: is there such a thing as a minetest "release cycle" (like a new minetest version all x months)… What determines the date of another stable engine version? 20:45 celeron55 it's "debian style" - it's released when it's ready 20:47 chmodsayshello but, I mean, there is a feature freeze? Is it just the coredevs saying "alright, we've got enoght features for a release"? 20:50 celeron55 a milestone for the next release is built up consisting of features and issues. it lives during development according to what seems possible and/or likely to happen "soon enough". once it's looking like a meaningful amount of features and fixes have been merged, then the remaining ones are merged or gotten off the list and then the feature freeze begins and release candidates are made 20:50 celeron55 feature freeze lasts until the release is finally made, and then the cycle repeats 20:51 chmodsayshello Oooh, yeah, now the roadmap approval I needed on a PR of mine makes sense, thanks! 20:52 celeron55 the core devs can't just make a release. at the least the feature freeze is to be had, but generally a milestone completion is seen through before that happens, in order to have a meaningful realease 20:52 celeron55 release* 20:53 celeron55 the roadmap approval is kind of a separate thing from the release cycle 20:56 celeron55 i wonder if we have a high level overview of that in place anywhere 20:58 celeron55 this is a bit complicated to describe on a whim... basically you have two options with a PR: either you do something that's on the pre-written roadmap (which is updated annually-ish), or you submit absolutely anything you like but a core dev has to individually approve the concept 20:58 chmodsayshello Actually, it seems like it: https://dev.minetest.net/Releasing_Minetest 20:59 chmodsayshello sorry, my bad for not finding that 20:59 celeron55 once the PR is concept-approved in either way, then it's subject to the normal review process 21:00 celeron55 see also these https://github.com/minetest/minetest/blob/master/.github/CONTRIBUTING.md https://github.com/minetest/minetest/blob/master/doc/direction.md 21:00 Krock even with the feature freeze there were bad releases in the past 8) 21:01 Krock happens 21:13 MTDiscord I am currently building the rendering engine 21:14 MTDiscord thats what i was trying to say 21:14 MTDiscord good luck! 21:14 MTDiscord Thanks. It's wgpu and SDL2, feels luxurious tbh 21:14 MTDiscord i wonder if this will have GOOD mobile support.. :juanchi_face: 21:15 MTDiscord mayhe not 21:15 MTDiscord doubt any of you would want to delve into the Andrew developement 21:16 MTDiscord Andrew development 21:16 MTDiscord Andrew 21:19 MTDiscord Yeah it has opengl es 21:23 MTDiscord neat 21:27 MinetestBot 02[git] 04appgurueu -> 03minetest/minetest: Fix revoke callbacks being run for `false` values passed to `set_priv… 13f0180ad https://github.com/minetest/minetest/commit/f0180ad488ec547758f56105b05e043db518086b (152024-01-22T21:24:32Z) 21:48 MinetestBot 02[git] 04appgurueu -> 03minetest/minetest: Minor documentation formatting fixes 13f6ecd93 https://github.com/minetest/minetest/commit/f6ecd931dc08a7a5a48b5d20787f6d8ba54462fc (152024-01-22T21:41:33Z)