Time Nick Message 02:37 cheapie Hmm.... I'm tempted to try to make a mod with a "rotor" entity that can spin (and slows down over time due to (simulated) friction), an "encoder" node that can detect the angular position of it and report over digilines, and some digilines-controllable (or mesecons or whatever) "electromagnet" nodes that can attract the rotor towards themselves, or not, depending on their input... 04:23 DeepThgt odd, the mods seem to be kept in memory when the server is running 04:23 DeepThgt but the conf sure isnt 05:03 DeepThgt ok, how can i "refresh" the node meta without digging the node? 05:04 DeepThgt i changed the formspec and inv slots on pipeworks autcrafters and if you toggle on off it only reflects the changed formspec geometry, not the slots 05:11 Baytuch| oh.. I want make mod about potatos. 05:13 DeepThgt what about potatoes? 05:13 DeepThgt just farming em? 05:13 DeepThgt or is there some deeper potato mechanics? 05:13 DeepThgt potato guns? 05:14 Baytuch| yeah, just farming, my friend 05:39 DeepThgt i still think potato guns would be a cool thing to add 06:40 Baytuch| https://youtu.be/foo-h7SjfqY 06:57 Baytuch| 00,04 ▶️  MineTest: user data on %APPDATA% (Windows x64) 13:43 __hex hi, how does the builtin mapgen prevent artifacts at chunk edges? 13:44 __hex emerge is giving me chunks that are slightly overlapping but it just overwrites one block with the other; am i supposed to write ignore to the outermost 8 node range? 13:49 calcul0n no, you should only fill the minp/maxp area 13:49 calcul0n what are you calling artifacts? 13:50 __hex what does "only fill" mean here though; i'm using mg_singlenode and on_generated 13:51 __hex mg_singlenode prefills the whole chunk with air 13:52 calcul0n i mean you should only generate the area between the minp/maxp parameters 13:53 calcul0n outside chunks have been or will be generated in another call of on_generated 13:53 __hex okay, harder question: what if i need to generate a bit around that area to make it work correctly 13:53 __hex like at least one node over minp-maxp for say, convolution or derivatives 13:54 __hex i guess i need to check what does the chunk contain before i write anything to it 13:54 calcul0n yep, you can check if node == ignore then write... 13:55 calcul0n but chunks are generated in arbitrary order, this will be the tricky part 13:56 __hex it's more like, i need to generate a bit around and then, in the final pass, zero it out back to what it was 13:57 calcul0n is your mapgen deterministic? in that case you could just generate the outside data but not write them 14:31 celeron55 __hex: i'm not sure what you need to do in your exact case, but the point of the extra is to be able to put e.g. a tree right on the edge so that it doesn't get cut in half with the other half forgotten 14:31 MTDiscord If you need to use the bleed area then you can, but it's best to design mapgen such that it doesn't need it. 14:32 MTDiscord You can't always reproduce a tree exactly the same because of randomness in some schematics, otherwise the bleed area probably wouldn't even be necessary. 14:33 celeron55 well the point simply is, you don't need to set up absolutely deterministic generation if you don't want to bother 14:33 celeron55 when you know you do have set it up, it could be helpful to be able to turn the edge buffer area off 14:34 celeron55 (in terms of performance) 14:34 MTDiscord Is that really an option? I didn't know it was even tunable... 14:34 celeron55 maybe it's still needed for lighting 14:35 celeron55 the exact behavior of the mapgen system is a decade old with no changes in this aspect, it's difficult to remember any reasonings for anything 14:36 celeron55 so what i recommend is: make changes and see what works :D 14:36 celeron55 if you find out something, post here so others can use the info 15:38 sfan5 __hex: make sure you leave the buffer area as 'ignore' in any case 15:38 sfan5 (except the parts you intentionally want to put nodes in) 15:40 __hex sfan5: thanks; that's the info i needed 15:42 sfan5 I'm pretty sure it should even be 'ignore' when you get handed the vmanip 17:26 MinetestBot 02[git] 04ZenonSeth -> 03minetest/minetest: Wireshark dissector: Made sure var 'pos' was only assigned locally to… 136783734 https://github.com/minetest/minetest/commit/6783734612832bd6a25adc3ddcd32f6a12c89249 (152023-11-23T17:26:00Z)