Time  Nick       Message
09:07 sfan5      unrealiable for what?
14:13 sfan5      rubenwardy: I changed more stuff so you'll have to take another look
14:46 rubenwardy currently the only game/mod that uses the return value of send_mapblock is https://github.com/AFCMS/Subway-Miner/blob/2090d143ec76a885762589ee7cf9cdadef04baa6/mods/sm_game/init.lua#L685
14:49 sfan5      is ITB open source? I remember sofar added it for that
14:50 rubenwardy https://gitlab.com/sofar/insidethebox/-/blob/master/mods/boxes/init.lua#L404
14:50 rubenwardy returning true for already sent would be fine for ItB
14:50 sfan5      so that code is actually wrong heh
14:51 rubenwardy *only game/mod on ContentDB
14:51 sfan5      searching github or gitlab for send_mapblock probably gives good results
14:52 sfan5      we shouldn't rely on the "all mods are on CDB" thing as much
14:52 rubenwardy searching on GH now, there's 13 pages but all so far don't capture the return value
14:52 rubenwardy CDB has a lot of mods, "doesn't break anything on CDB" is a good enough metric when changing undefined behaviour
14:53 rubenwardy ok, the only one on GitHub that captures the return value is Subway Miner
14:55 rubenwardy AFCMS: too tired to work out what subway miner is doing with infos.is_send, what did you expect send_mapblock to return when the client already has the mapblock?
14:55 MTDiscord  <rubenwardy> @AFCM
17:07 erle       rubenwardy are you searching mesehub too?
17:08 rubenwardy I'm not going to search every single git repository host that has Minetest mod
17:25 erle       rubenwardy i am asking mostly because a big chunk of the mcl* mods are hosted there
17:25 erle       but ig most of those are on cdb
17:28 erle       sfan5, i have noticed lbms to become unreliable when replacing large amounts of nodes. cora has tried to use them for “structure blocks“ (basically single-node-decorations that are supposed to “unpack” a schematic) and found them unreliable enough that she is trying all kinds of workarounds.
17:29 erle       apparently they are not unreliable when replacing large amounts of nodes, they are just unreliable – and wanting to replace lots of stuff makes it obvious.
17:29 erle       i believe warr1024 knows more about it than i do
17:30 rubenwardy lbms run when mapblocks load, not when replacing nodes
17:31 erle       i can tell you one case where i ran into it: mineclone <0.71 had no air above the nether, just void. i made an lbm that created 128 nodes of air there, so you can build on the nether roof (a bedrock plains).
17:33 erle       rubenwardy does that make it more clear what i was trying to achieve?
17:34 rubenwardy changing the map during an lbm causes the new nodes to be missed?
17:34 rubenwardy eh
17:34 erle       eh, i think i should make a proof-of-concept to show it
17:37 erle       rubenwardy something like this, you see any issue with it? https://git.minetest.land/Mineclonia/Mineclonia/commit/5f5796330f89e42eea07fbb79ec69c4017478160
17:37 erle       it's extremely laggy at replacing stuff too
17:42 MTDiscord  <luatic> erle: Have you verified that the issue is actually with the ABMs and not with the node placement?
17:42 MTDiscord  <luatic> LBMs*
17:43 erle       luatic tell me more
17:45 Krock      will merge #12421 #11931 #11846  and #11427  in 15 minutes
17:45 ShadowBot  https://github.com/minetest/minetest/issues/12421 -- Inline triLinearInterpolationNoEase and triLinearInterpolation by paradust7
17:45 ShadowBot  https://github.com/minetest/minetest/issues/11931 -- on_deactivate: distinguish removal and unloading by appgurueu
17:45 ShadowBot  https://github.com/minetest/minetest/issues/11846 -- No damage effects on hp_max change by appgurueu
17:45 ShadowBot  https://github.com/minetest/minetest/issues/11427 -- Redo serialize.lua by appgurueu
17:49 MTDiscord  <luatic> erle: I don't have more tell, I'm just asking whether you've isolated the issue of "LBM doesn't trigger when it should" or rather the issue of "LBM that is supposed to place blocks fails to do so" which may be either related to block placement bugs or LBM bugs
17:50 erle       you are definitely on to something here. namely, i have not added debug statements to the lbm to figure out when it is run.
17:59 Krock      merging
18:01 Krock      done
22:43 Zughy[m]   I think #12429 should be high priority for the next release, it's free x-ray everywhere
22:43 ShadowBot  https://github.com/minetest/minetest/issues/12429 -- Seeing through blocks glitch