Time |
Nick |
Message |
00:19 |
|
yang2003 joined #minetest-dev |
00:37 |
|
Void7 joined #minetest-dev |
01:01 |
sofar |
ok, either I'm going nuts or there's something really wrong with getIds() |
01:03 |
sofar |
nope, that's not it either |
01:05 |
sofar |
if (f1.connects_to_ids.find(to.param0) == f1.connects_to_ids.end()) return; |
01:05 |
sofar |
why does it claim the id is listed there? |
01:09 |
sofar |
I think my map is getting corrupted o_O |
01:21 |
|
Void7 joined #minetest-dev |
01:26 |
|
Void7 left #minetest-dev |
01:30 |
sofar |
ohhhhh finally, finally I find the cause of that darn bug |
02:29 |
|
Dragonop joined #minetest-dev |
02:38 |
|
paramat joined #minetest-dev |
02:42 |
paramat |
!tell celeron55 i'm still working on mgwatershed https://github.com/paramat/minetest/tree/rivergen adding volcanos, multiple floatland layers and vertical megastructures to get up there. will be perfect for farmap |
02:42 |
ShadowBot |
paramat: O.K. |
03:04 |
paramat |
can anyone +1 #3851 ? it's almost trivial but i'd rather have another approval |
03:04 |
ShadowBot |
https://github.com/minetest/minetest/issues/3851 -- Mapgen: Unhide singlenode mapgen by paramat |
03:41 |
|
Anchakor joined #minetest-dev |
04:50 |
|
Miner_48er joined #minetest-dev |
05:40 |
hmmmm |
sometimes PR approval isn't always about the code itself, but rather if the feature itself is wanted |
05:53 |
|
neoascetic joined #minetest-dev |
06:05 |
paramat |
hi hmmmm possible issue with a unit test failing on good PRs #3786 |
06:05 |
ShadowBot |
https://github.com/minetest/minetest/issues/3786 -- Unreliably failing unittests on mac |
06:05 |
|
est31 joined #minetest-dev |
06:05 |
hmmmm |
well, i would fix it if i had a working mac os x setup |
06:05 |
paramat |
oh ok |
06:06 |
est31 |
there is a PR around which tries to fix it |
06:06 |
est31 |
its referenced in the issue |
06:08 |
hmmmm |
est31, what do you think about #3851? |
06:08 |
ShadowBot |
https://github.com/minetest/minetest/issues/3851 -- Mapgen: Unhide singlenode mapgen by paramat |
06:08 |
hmmmm |
do you agree with it? |
06:08 |
hmmmm |
should singlenode be selectable by the end user? |
06:08 |
est31 |
I see users be in favor for it |
06:08 |
hmmmm |
but what's your personal opinion on the matter |
06:09 |
est31 |
I don't have an use case for me |
06:09 |
est31 |
never needed it |
06:09 |
est31 |
but I see the users want it to e.g. create a world for worldedit etc |
06:09 |
est31 |
idk, it _can_ be done with mods too |
06:09 |
est31 |
it is very bad though when its in the default menu |
06:10 |
est31 |
and sb tries to select that option in order to find out what its doing |
06:10 |
est31 |
they will think its broken |
06:10 |
est31 |
or worse, minetest |
06:10 |
est31 |
perhaps have a setting that activates it? |
06:11 |
hmmmm |
that's silly |
06:11 |
paramat |
nah, lets either hide or not hide |
06:11 |
paramat |
i guess i'm not too bothered, i just get asked a lot about it |
06:11 |
hmmmm |
one small issue |
06:11 |
est31 |
well the issue is that its a simple drop down menu |
06:12 |
est31 |
no description of the mapgens |
06:12 |
hmmmm |
because of the way the main menu config works, you can't manually type "singlenode" in the config file |
06:12 |
hmmmm |
because it would show up as blank in the dropdown box |
06:13 |
hmmmm |
that would be the ideal compromise, IMHO, have the setting configurable by the text .conf, but not selectable by the GUI menu |
06:13 |
hmmmm |
power users shouldn't be left out in the cold |
06:13 |
|
Hunterz joined #minetest-dev |
06:13 |
hmmmm |
but then again, at the same time, I can't help but argue that there are too many .conf options, and that many of them should be removed and instead replaced with a mod API |
06:13 |
est31 |
the issue currently is that you can't select/deselect mods at world creation, right? |
06:14 |
hmmmm |
because really if you think about it |
06:14 |
est31 |
otherwise you could write a "singlenode activation" mod that does nothing except that |
06:14 |
hmmmm |
isn't coding basic Lua almost the same as writing a .conf file? |
06:14 |
est31 |
and then the user simply activates the mod at world creation |
06:14 |
hmmmm |
right and there are other problems with that too |
06:14 |
est31 |
it sort of is yeah |
06:15 |
est31 |
minetest.register_node is mostly just passing a big table to the server |
06:15 |
hmmmm |
right and that's exactly what Lua was intended for |
06:15 |
paramat |
i've noticed that the code in default/mapgen.lua that selects ore/biome/deco registrations and nyancat generation on chosen mapgen, now fails for singlenode since it was hidden. so we now get nyancats in singlenode when they're meant to be disabled https://github.com/minetest/minetest_game/blob/master/mods/default/mapgen.lua#L1409 |
06:16 |
hmmmm |
celeron didn't go forth with this mission to create an engine with games defined by mods |
06:16 |
hmmmm |
it was just to add another layer of customization |
06:16 |
hmmmm |
a glorified .conf file for bulk data |
06:16 |
est31 |
minetest is not a real engine yeah |
06:17 |
est31 |
we have this restriction that we can't grant full opengl access on the client for example |
06:17 |
est31 |
not even shaders |
06:17 |
hmmmm |
paramat: did you actually notice it was hidden, or do you just think that code is probably faulty? |
06:19 |
paramat |
well default will run before a mod that can select singlenode, so it makes sense i guess |
06:19 |
hmmmm |
maybe on_mapgen_init needs to make a comeback |
06:19 |
est31 |
and we need mainmenu changes |
06:19 |
paramat |
i tried adding that back in, it didn't help |
06:19 |
est31 |
err |
06:19 |
est31 |
sorry no we dont need |
06:20 |
est31 |
just realized its the "create world" button |
06:20 |
est31 |
not the "create world and play" button |
06:21 |
est31 |
what did on_mapgen_init do? |
06:22 |
hmmmm |
it was a callback that happened after all mods initially ran, most things were registered, but before the first environment step happened |
06:24 |
est31 |
would we need it for selecting singlenode? |
06:24 |
est31 |
we can just modify the setting at load time, no? |
06:26 |
hmmmm |
no, the problem is that default runs before most other mods, and it needs to make a decision on which decorations to register based on what the mapgen will be |
06:26 |
hmmmm |
simply put, default cannot predict the future |
06:27 |
est31 |
yeah |
06:27 |
est31 |
idk how all that mapgen stuff works |
06:27 |
hmmmm |
i'll tell you how it works |
06:27 |
est31 |
is it harmful to have decorations in singlenode? |
06:27 |
paramat |
they don't appear |
06:28 |
est31 |
it shouldn't invoke the decorations system so to speak. no? |
06:28 |
paramat |
stuff just gets registered |
06:28 |
hmmmm |
we have all these mods that want to change their behavior based on the final state of game initialization, but mods are loaded in a certain order |
06:28 |
paramat |
the only problem is nyancats heh |
06:28 |
paramat |
(minor problem) |
06:29 |
hmmmm |
paramat, a workaround is to check for the mapgen in the on_generated callback |
06:29 |
paramat |
aha |
06:29 |
paramat |
will do that |
06:30 |
est31 |
but nyan cats are only a problem on the first run, right? |
06:30 |
est31 |
after that the mapgen gets set to singlenode |
06:30 |
hmmmm |
make a global for mapgen.lua called g_effective_mapgen, and in the first on_generated callback, if it's nil, set it to minetest.get_mapgen_params().mgname |
06:30 |
hmmmm |
if g_effective_mapgen == "singlenode" then return end |
06:31 |
est31 |
or local even |
06:31 |
hmmmm |
or g_is_mapgen_singlenode |
06:31 |
hmmmm |
cache it |
06:31 |
paramat |
ok |
06:31 |
est31 |
it should be local to the file |
06:31 |
hmmmm |
doing a heafty string comparison op every single callback is not the end of the world by any means, but a totally unnecessary that's trivial to solve |
06:31 |
hmmmm |
totally unnecesary one* |
06:32 |
paramat |
#3850 |
06:32 |
ShadowBot |
https://github.com/minetest/minetest/issues/3850 -- Mapgen: Fix light in tunnels at mapchunk borders by paramat |
06:33 |
hmmmm |
ehh |
06:34 |
hmmmm |
i appreciate the simplicity, but creating a flat ceiling for caves is nasty |
06:34 |
hmmmm |
this is almost as bad as the walled-in-cave problem |
06:34 |
paramat |
nooo the 'roof' gets excavated when the upper chunk is generated |
06:35 |
paramat |
tunnels are unnaffected, just not lit up at every border |
06:35 |
hmmmm |
it doesn't. different tunnels get carved in different chunks, due to the differing blockseeds |
06:35 |
hmmmm |
unless this only applies to noise-based caves |
06:35 |
paramat |
these are the 3D noise tunnels |
06:35 |
hmmmm |
why are they carving outside of the chunk borders at all, then?? |
06:36 |
hmmmm |
that's wrong |
06:36 |
paramat |
1 up 1 down overgeneration in all mapgens other than mgv6 |
06:36 |
paramat |
yes it was unnecessary |
06:36 |
hmmmm |
so in other words this is your bug |
06:36 |
paramat |
yes |
06:37 |
hmmmm |
ugh alright fine |
06:37 |
hmmmm |
approved |
06:37 |
paramat |
ok thanks |
06:38 |
paramat |
the pseudorandom caves above water level still get lit up at chunk borders, we could perhaps consider disabling them above water level |
06:40 |
paramat |
(except in mgv6 of course) |
07:08 |
|
Obani joined #minetest-dev |
07:12 |
|
Ritchie_ joined #minetest-dev |
07:13 |
|
Taoki joined #minetest-dev |
07:14 |
|
sfan5_ joined #minetest-dev |
07:15 |
|
cat5e joined #minetest-dev |
07:16 |
|
misprint_ joined #minetest-dev |
07:16 |
|
Sokomine joined #minetest-dev |
07:17 |
paramat |
ok i've changed my mind, i think singlenode should remain hidden |
07:19 |
Obani |
I don't think so |
07:19 |
Obani |
Modders cannot always compile their Minetest |
07:23 |
paramat |
to select singlenode you only need a simple mod |
07:23 |
paramat |
this: https://gist.github.com/paramat/feb2361da106371652ad |
07:24 |
paramat |
it doesn't even need 'register on mapgen init' anymore |
07:25 |
Obani |
paramat, singlenode is useful for lua mapgens isn't it ? |
07:25 |
paramat |
yep, those will already auto-select it |
07:31 |
paramat |
nore sfan5_ 3 simple reviews game#922 game#913 game#758 |
07:31 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/922 -- Fencegate: prevent y-fighting at top. by sofar |
07:31 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/913 -- Fire: Optimize spreading ABM. by sofar |
07:31 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/758 -- Add sounds for steel doors. by sofar |
07:32 |
|
Taoki joined #minetest-dev |
07:37 |
|
Calinou joined #minetest-dev |
07:47 |
est31 |
opinions on #3854 ? |
07:47 |
ShadowBot |
https://github.com/minetest/minetest/issues/3854 -- Drop support for < 25 protocol servers by default |
07:51 |
|
Taoki joined #minetest-dev |
07:54 |
paramat |
looking |
07:57 |
est31 |
hmmmm, perhaps interesting for you too |
07:57 |
est31 |
only about your opinion |
08:05 |
|
Megaf joined #minetest-dev |
08:12 |
|
neoascetic joined #minetest-dev |
08:12 |
|
neoascetic joined #minetest-dev |
08:17 |
|
Taoki joined #minetest-dev |
08:17 |
|
technics joined #minetest-dev |
08:22 |
paramat |
will merge #3535 #3850 soon |
08:22 |
ShadowBot |
https://github.com/minetest/minetest/issues/3535 -- fix #2876 (fix player teleportation bug) by HybridDog |
08:22 |
ShadowBot |
https://github.com/minetest/minetest/issues/3850 -- Mapgen: Fix light in tunnels at mapchunk borders by paramat |
08:22 |
est31 |
fine |
08:30 |
|
neoascetic joined #minetest-dev |
08:35 |
|
Taoki joined #minetest-dev |
08:41 |
paramat |
now merging 3535 3850 |
08:50 |
paramat |
merged |
08:51 |
|
Taoki joined #minetest-dev |
08:54 |
paramat |
wow under 120 PRs now |
09:00 |
celeron55 |
i always look at it as the ratio of closed/open PRs |
09:00 |
celeron55 |
it's now >16 |
09:17 |
|
Megaf joined #minetest-dev |
09:21 |
|
linkedinyou joined #minetest-dev |
09:25 |
|
nrzkt joined #minetest-dev |
09:31 |
|
DevBox joined #minetest-dev |
10:41 |
|
paramat joined #minetest-dev |
10:51 |
|
Samson1 joined #minetest-dev |
10:51 |
|
Dragonop joined #minetest-dev |
11:18 |
|
proller joined #minetest-dev |
11:21 |
|
Taoki joined #minetest-dev |
11:39 |
|
Calinou joined #minetest-dev |
11:51 |
|
proller joined #minetest-dev |
11:57 |
|
proller joined #minetest-dev |
11:59 |
|
proller joined #minetest-dev |
12:04 |
|
proller joined #minetest-dev |
12:47 |
|
DFeniks joined #minetest-dev |
13:07 |
|
Player_2 joined #minetest-dev |
13:08 |
|
Fixer joined #minetest-dev |
13:10 |
|
Darcidride joined #minetest-dev |
13:15 |
|
STHGOM joined #minetest-dev |
13:20 |
|
neoascetic joined #minetest-dev |
13:30 |
|
kahrl joined #minetest-dev |
13:34 |
|
rubenwardy joined #minetest-dev |
13:54 |
|
neoascetic joined #minetest-dev |
13:57 |
|
rubenwardy joined #minetest-dev |
14:04 |
|
Darcidride_ joined #minetest-dev |
14:23 |
|
rubenwardy_ joined #minetest-dev |
14:34 |
|
STHGOM joined #minetest-dev |
14:40 |
|
celeron55 joined #minetest-dev |
14:56 |
|
kaadmy joined #minetest-dev |
15:02 |
|
hmmmm joined #minetest-dev |
15:12 |
|
damiel joined #minetest-dev |
15:14 |
|
hmmmmm joined #minetest-dev |
15:17 |
|
Void7 joined #minetest-dev |
15:17 |
|
Taoki[mobile] joined #minetest-dev |
15:41 |
|
proller joined #minetest-dev |
15:41 |
Fixer |
with default farmap settings managed to reach 1.6gb of ram usage with constant exploring on mapgenv6 |
15:53 |
|
proller joined #minetest-dev |
15:55 |
|
Hunterz joined #minetest-dev |
16:10 |
|
Obani joined #minetest-dev |
16:41 |
|
Krock joined #minetest-dev |
17:08 |
|
proller joined #minetest-dev |
17:16 |
|
darkrose joined #minetest-dev |
17:16 |
|
darkrose left #minetest-dev |
17:19 |
|
blaze joined #minetest-dev |
17:37 |
Fixer |
2.2gb on NYC map |
17:38 |
Fixer |
lua out of memory error after some time when "far_map_allow_generate = true" |
17:57 |
|
rubenwardy_ joined #minetest-dev |
17:58 |
|
Calinou joined #minetest-dev |
17:58 |
|
damiel joined #minetest-dev |
18:00 |
|
Siva joined #minetest-dev |
18:19 |
|
bugzapper joined #minetest-dev |
18:34 |
|
SaKeLWorld joined #minetest-dev |
18:58 |
|
rubenwardy_ joined #minetest-dev |
19:03 |
|
Amaz joined #minetest-dev |
19:07 |
|
Void7 joined #minetest-dev |
19:40 |
|
jai joined #minetest-dev |
19:49 |
|
Guest44693 joined #minetest-dev |
19:53 |
|
ud1 joined #minetest-dev |
20:06 |
|
proller joined #minetest-dev |
20:46 |
|
Megaf joined #minetest-dev |
20:51 |
|
SaKeLWorld joined #minetest-dev |
21:05 |
|
ud1__ joined #minetest-dev |
21:42 |
|
dfelinto joined #minetest-dev |
21:54 |
|
kaeza joined #minetest-dev |
22:07 |
|
Miner_48er joined #minetest-dev |
22:09 |
|
proller joined #minetest-dev |
23:09 |
|
Void7 joined #minetest-dev |
23:48 |
|
damiel joined #minetest-dev |