| Time |
Nick |
Message |
| 03:32 |
|
vampirefrog joined #luanti |
| 03:45 |
|
amfl joined #luanti |
| 04:37 |
|
aliasreadytaken joined #luanti |
| 05:00 |
|
MTDiscord joined #luanti |
| 05:58 |
|
rubenwardy joined #luanti |
| 06:00 |
|
burgeresque joined #luanti |
| 07:18 |
|
FeXoR joined #luanti |
| 07:31 |
mrcheese |
!puzzle |
| 07:31 |
mrcheese |
wait wrong channel |
| 07:31 |
mrcheese |
xd |
| 09:14 |
whosit |
is there an API for checking when code is evaluated? like during mod load time, or when game is running... |
| 09:15 |
whosit |
ah, there's INIT, but it's not documented? |
| 09:15 |
whosit |
so, should not be used? |
| 09:16 |
whosit |
also, does not allow to distinguish mod load time... |
| 09:22 |
SwissalpS |
what is your goal whosit? maybe there is another way to achieve what you are looking for. |
| 09:25 |
SwissalpS |
when a mod is being loaded its init.lua is run, you can set a flag there and then either with a core.after() event or core.register_on_mods_loaded() clear the flag again. |
| 09:25 |
whosit |
yeah, I usually have to have mod.initialized flag... |
| 09:26 |
whosit |
I have /reload_file command, wanted to conditionally execute part of file (registration) during mod load, but not when I manually /reload_file it... |
| 09:26 |
whosit |
I guess `if initialized` is basically same :p |
| 09:27 |
SwissalpS |
you can't execute any /<command> unless everything is loaded....afaik |
| 09:28 |
whosit |
it's a separate snippet for reloading arbitrary files... |
| 09:28 |
SwissalpS |
sounds like you only need a local variable 'first_run' in your mod. |
| 09:28 |
whosit |
yes, but I'm in a mood to do things in weird ways :p |
| 09:28 |
SwissalpS |
good luck :D |
| 09:33 |
whosit |
still, there must be a way (and having an official would not hurt :p) |
| 09:34 |
MinetestBot |
[git] Emojigit -> luanti-org/luanti: Add support for JPEG mod screenshots (#16911) 6dd22dd https://github.com/luanti-org/luanti/commit/6dd22ddac3a0997156697033febd6764bf1ac57b (2026-02-06T09:34:34Z) |
| 09:34 |
MinetestBot |
[git] appgurueu -> luanti-org/luanti: Document different visual size default for players 08f9439 https://github.com/luanti-org/luanti/commit/08f943902e4342628de4c0ca9e59ef8915744638 (2026-02-06T09:34:43Z) |
| 09:34 |
MinetestBot |
[git] sfan5 -> luanti-org/luanti: Report fatal client startup errors via message box ef36023 https://github.com/luanti-org/luanti/commit/ef360231707558e318c7989e3c52c9879fd79377 (2026-02-06T09:34:52Z) |
| 09:34 |
MinetestBot |
[git] sfan5 -> luanti-org/luanti: Handle unified GL driver init errors correctly b1ed8f2 https://github.com/luanti-org/luanti/commit/b1ed8f25cefbf96e29519fc516f2f92b64ed4d24 (2026-02-06T09:34:52Z) |
| 10:47 |
|
Juri joined #luanti |
| 10:47 |
|
Niklp joined #luanti |
| 12:13 |
|
sfan5 joined #luanti |
| 12:14 |
|
mrkubax10 joined #luanti |
| 12:20 |
|
ireallyhateirc joined #luanti |
| 12:51 |
|
kimapr joined #luanti |
| 13:37 |
|
user333_ joined #luanti |
| 13:38 |
|
user333_8 joined #luanti |
| 13:42 |
|
user333_2 joined #luanti |
| 13:49 |
|
PoochInquisitor joined #luanti |
| 13:50 |
|
user333_2 joined #luanti |
| 14:03 |
|
jaca122 joined #luanti |
| 14:16 |
ireallyhateirc |
Are block updates for far away blocks less frequent? |
| 14:17 |
[MatrxMT] |
<Blockhead256> anything that's actually loaded should be running ABMs at the same rate, but clients can easily render further than the simulation |
| 14:18 |
ireallyhateirc |
no I don't mean the active block area, I mean all loaded blocks |
| 14:18 |
ireallyhateirc |
up to the horizon |
| 14:19 |
celeron55 |
the set of loaded blocks is different for the server vs. the client |
| 14:19 |
[MatrxMT] |
<Blockhead256> the only ones that can be loaded by the server are within the server settings for client view distance and the forceloads |
| 14:20 |
[MatrxMT] |
<Blockhead256> when you can see a long way away on a client, most of that isn't being simulated, only what's close |
| 14:20 |
sfan5 |
you might be seeing block updates to the client come in later |
| 14:20 |
[MatrxMT] |
<Blockhead256> with the zoom-in control, you can load terrain that's far away |
| 14:21 |
ireallyhateirc |
my problem is that I'm writing a mod which modifies blocks with VMs and I want all loaded blocks to be changed and visible to the client |
| 14:21 |
celeron55 |
yeah and the game obviously doesn't stop to synchronize blocks between the client and the server, ever. block updates are always done in a best-effort manner |
| 14:22 |
ireallyhateirc |
but it seems that far away blocks get updated with delay |
| 14:22 |
celeron55 |
s/game/engine/ |
| 14:23 |
sfan5 |
you can use core.send_mapblock |
| 14:23 |
ireallyhateirc |
okay, I'll try this |
| 14:29 |
ireallyhateirc |
I thought I was playing 5d chess but it was checkers against a pigeon |
| 14:34 |
ireallyhateirc |
sfan5, I think you mean player:send_mapblock ? |
| 14:35 |
ireallyhateirc |
yeah seems like there's no core.send_mapblock |
| 14:38 |
sfan5 |
i guess |
| 14:51 |
ireallyhateirc |
could it be that blocks emerge, get cached by the client and then get unloaded? |
| 14:52 |
celeron55 |
of course |
| 14:54 |
ireallyhateirc |
this is what I see: https://files.catbox.moe/q4v8lc.png |
| 14:54 |
ireallyhateirc |
dark soil, close to the player: correctly changed |
| 14:55 |
ireallyhateirc |
bright soil, far away: not changed |
| 14:56 |
ireallyhateirc |
to detect loaded/active blocks I use my AI slop branch for mapblock callbacks: https://github.com/luanti-org/luanti/pull/16910 |
| 14:56 |
ireallyhateirc |
VMs are spawned for each mapblock that just got loaded |
| 15:09 |
cheapie |
I'm sure there's something I'm supposed to be looking at with the map but I'm too distracted by whatever's going on with that hand |
| 15:16 |
[MatrxMT] |
<Blockhead256> that hand looks amazing |
| 15:22 |
ireallyhateirc |
Blockhead256, you mean the block callbacks thing or the unfinished landscape? |
| 15:25 |
[MatrxMT] |
<Blockhead256> the picture o the landscape with a very high res hand |
| 15:48 |
ireallyhateirc |
is it possible that blocks could emerge, get sent to client but not be loaded? |
| 15:48 |
ireallyhateirc |
as in never loaded |
| 15:56 |
sfan5 |
only loaded blocks can be sent to the client |
| 16:03 |
ireallyhateirc |
I'll recheck my AI slop code then, maybe it actually doesn't fetch all loaded blocks |
| 16:56 |
rubenwardy |
We gave out 1,500 Luanti stickers, 300 Lua stickers, and 200 Luanti pride stickers at FOSDEM |
| 16:57 |
ireallyhateirc |
we had a FOSDEM again? great |
| 16:58 |
ireallyhateirc |
I'm too lazy/greedy to attend but are people having fun there? |
| 17:13 |
|
crazylad joined #luanti |
| 17:17 |
MTDiscord |
<the4spaceconstants2181> "stingy" may work in place of "greedy", or "cheap" |
| 18:01 |
|
user333_ joined #luanti |
| 18:01 |
|
user333_2 joined #luanti |
| 18:02 |
|
user333_ joined #luanti |
| 18:06 |
|
user333_8 joined #luanti |
| 18:06 |
|
user333_8 joined #luanti |
| 18:07 |
|
jaca122 joined #luanti |
| 18:07 |
|
Guest389 joined #luanti |
| 18:18 |
|
Guest389 joined #luanti |
| 18:20 |
|
user333_ joined #luanti |
| 18:24 |
|
user333_ joined #luanti |
| 18:24 |
|
user333_ joined #luanti |
| 18:27 |
|
mrcheese joined #luanti |
| 18:33 |
|
mrcheese joined #luanti |
| 18:55 |
|
qqq joined #luanti |
| 19:06 |
whosit |
sometimes static_save=true entities go "missing": they don't load with mapblock, but after mapblock is unloaded and loaded again, they appear where they should be. |
| 19:07 |
whosit |
how to debug this?.. |
| 19:17 |
celeron55 |
i would augmentate the mapblock saving and loading code to log the entity data, and then go from there, trying to track what's happening in the log - maybe filter by specific data to get rid of noise. the most important thing is to build a reproducible test case. figure out how to make it happen consistently. if you can do that, then you should post an issue on github because i bet many people are |
| 19:17 |
celeron55 |
interested in figuring it out |
| 19:24 |
whosit |
well, it's a static entity that loads/unloads many times per day on a server, it can disappear like once per week |
| 19:24 |
whosit |
I don't even know what can be logged about it on lua side... |
| 19:26 |
whosit |
I know for sure that reloading block brings it back, because while it was missing, I placed a "canary" there. It printed a message in on_deactivate, and when I came back - missing entity was there |
| 19:27 |
whosit |
when it's missing, it's just not anywhere either (and if was moved, it would not appear back where it should be) |
| 19:28 |
whosit |
I guess adding some debug to C side would help, but kinda not ideal on an active server :p |
| 19:29 |
whosit |
people notice it because it's a missing NPC, but I think different entities from different mods were missing too... |
| 19:34 |
whosit |
I think that it's missing while mapblock is active, but reappears on reactivation should be a big hint, but I don't know luanti internals ;p |
| 19:35 |
whosit |
I see "failed to create active object from static object" in the source... |
| 19:35 |
whosit |
I'll ask to check logs for that... |
| 19:36 |
sfan5 |
i mean that sounds like cause of "entity doesn't reappear" |
| 19:37 |
whosit |
" If couldn't create object, store static data back." |
| 19:37 |
whosit |
https://github.com/luanti-org/luanti/blob/b1ed8f25cefbf96e29519fc516f2f92b64ed4d24/src/serverenvironment.cpp#L1595 |
| 19:58 |
whosit |
corrupted data would not make sense... since it would not be able to just load next time normally... |
| 20:20 |
|
mrkubax10 joined #luanti |
| 23:13 |
|
mrkubax10 joined #luanti |
| 23:33 |
|
panwolfram joined #luanti |
| 23:43 |
user333_ |
i'm open-sourcing my PVP game in ~15min, after that you can insult me for the terrible codebase |
| 23:45 |
sfan5 |
is it slop-free? |
| 23:45 |
user333_ |
yes |
| 23:45 |
user333_ |
i wrote a lot of it, two friends contributed as well |
| 23:48 |
mrcheese |
i've constantly advised we burn the codebase down yet it hasnt happened xd |
| 23:50 |
user333_ |
"who needs namespacing if we can just make everything a global?" |
| 23:50 |
user333_ |
> ^^^ how it was partially written |
| 23:50 |
mrcheese |
*internal screaming* |
| 23:56 |
user333_ |
okay, here we go... |
| 23:57 |
user333_ |
done |
| 23:58 |
user333_ |
https://github.com/a-bad-dev/simple-shooter-game |
| 23:58 |
ireallyhateirc |
Speaking of slop, my PR is ready #16910 |
| 23:58 |
ShadowBot |
https://github.com/luanti-org/luanti/issues/16910 -- Add mapblock callbacks and tracking tables by kromka-chleba |
