Time |
Nick |
Message |
00:11 |
|
luk3yx joined #minetest-dev |
00:25 |
|
ShadowBot joined #minetest-dev |
01:01 |
|
jordan4ibanez joined #minetest-dev |
01:46 |
|
v-rob joined #minetest-dev |
01:54 |
|
tekakutli joined #minetest-dev |
02:04 |
|
Soni joined #minetest-dev |
02:55 |
v-rob |
Why does minetest.encode_png() say that string.char() is unsafe for binary data? Lua strings are ASCII, and `assert(i == string.byte(string.char(i)))` never fails for numbers in the range [0, 255] (tested on both LuaJIT and standard Lua) |
02:56 |
v-rob |
I'm sure plenty of Lua programs rely on string.char()/string.byte() working for purely binary data. |
03:13 |
|
hmmmm joined #minetest-dev |
03:14 |
|
v-rob joined #minetest-dev |
04:00 |
|
MTDiscord joined #minetest-dev |
04:07 |
|
YuGiOhJCJ joined #minetest-dev |
04:47 |
|
BuckarooBanzai joined #minetest-dev |
04:56 |
|
BuckarooBanzai joined #minetest-dev |
05:28 |
|
imi joined #minetest-dev |
05:30 |
|
calcul0n joined #minetest-dev |
05:46 |
|
calcul0n joined #minetest-dev |
05:49 |
|
tekakutli joined #minetest-dev |
06:50 |
|
appguru joined #minetest-dev |
07:22 |
sfan5 |
hecks said he had issues with it in testing, I also mentioned this during review but we ultimately left it in |
07:56 |
MTDiscord |
<luatic> yeah, I couldn't reproduce this either, Lua is supposed to work with byte strings and uses (unsigned) char in the code so should be fine? |
07:57 |
MTDiscord |
<luatic> not sure which obscure platforms hecks tested this on though |
07:57 |
|
YuGiOhJCJ joined #minetest-dev |
08:14 |
sfan5 |
windows :^) |
08:19 |
jordan4ibanez[m] |
That's pretty obscure |
10:08 |
celeron55 |
i agree with GreenXenith. however annoying it is that projects use names that have nothing to do with anything at all, there's that reason for it. given years or decades, about half of the time projects evolve outside of the original name |
10:42 |
|
appguru joined #minetest-dev |
11:25 |
|
proller joined #minetest-dev |
11:26 |
|
appguru joined #minetest-dev |
11:40 |
|
appguru joined #minetest-dev |
12:12 |
|
appguru joined #minetest-dev |
12:19 |
|
proller joined #minetest-dev |
12:55 |
|
Noisytoot joined #minetest-dev |
13:44 |
|
pgimeno joined #minetest-dev |
15:20 |
MTDiscord |
<PrairieWind> how hard would it be for me to try to continue #12907 ? |
15:20 |
ShadowBot |
https://github.com/minetest/minetest/issues/12907 -- Add paramtype2 colorflowingliquid by TurkeyMcMac |
15:42 |
|
Desour joined #minetest-dev |
15:49 |
rubenwardy |
How does the engine optimise node boxes? |
15:52 |
Desour |
I recently noticed that the sides that are next to solid nodes are not drawn. not sure if this is new. |
15:52 |
rubenwardy |
Most importantly, how do they vary from mesh nodes? |
15:52 |
rubenwardy |
I remember that hecks(?) did some optimisations last year, and something about node boxes being converted to meshes anyway |
15:57 |
rubenwardy |
also, is a node box a single cuboid or the sum of all cuboids |
16:02 |
Desour |
there's #12262 and #10122. #13216 may also have done some changes |
16:02 |
ShadowBot |
https://github.com/minetest/minetest/issues/12262 -- Avoid rendering invisible faces of simple nodeboxes by x2048 |
16:02 |
ShadowBot |
https://github.com/minetest/minetest/issues/10122 -- Render nodeboxes + meshes with opaque material if possible by sfan5 |
16:02 |
ShadowBot |
https://github.com/minetest/minetest/issues/13216 -- Remove fast faces by numberZero |
16:06 |
Desour |
(and it's all the cuboids, see NodeBox type) |
16:07 |
rubenwardy |
argh |
16:08 |
rubenwardy |
should be NodeBoxes |
16:08 |
rubenwardy |
I swear, Minetest defies explanation |
16:09 |
Desour |
nodebox = nodeboks = nodebokss = nodeboxs = nodeboxes :D |
16:09 |
|
fluxionary joined #minetest-dev |
16:10 |
rubenwardy |
I had a professor that spelt it boxs because he didn't like that es was two more characters |
16:10 |
rubenwardy |
this was my haskell professor |
16:12 |
MTDiscord |
<PrairieWind> english is evil sometimes. |
16:14 |
rubenwardy |
ChatGPT: NodeBox is a feature in Minetest, an open-source sandbox video game, that allows players to create 3D models and shapes out of small cubes called nodes. A NodeBox is essentially a virtual 3D container or box in which nodes can be arranged to create shapes and objects. It is a powerful tool for creating custom objects, buildings, and landscapes in the game. |
16:15 |
Desour |
I find it interesting how collision code seems to use drawtype and f.node_box (as opposed to f.collision_box) |
16:17 |
Desour |
I'm starting to get the feeling ChatGPT is just a tool to trigger cunningham's law |
16:19 |
MTDiscord |
<Warr1024> It's supposed to be a tool for getting answers more efficiently, so why wouldn't it take advantage of that? |
16:20 |
Desour |
oh wow. so the correct way to use it is not to ask it questions and believe the answers, but to copy the answers to some forum or chat, and then to wait for corrections? amazing idea! |
16:26 |
MTDiscord |
<PrairieWind> hmm, lets try it. |
16:26 |
MTDiscord |
<PrairieWind> Minetest is a game. |
16:39 |
|
dv^_^ joined #minetest-dev |
16:40 |
MTDiscord |
<Flamore> Minetest game is a game mhm |
16:47 |
MTDiscord |
<luatic> Yeah. People are replacing the "here's what dumb I did" in their questions with "here's what dumb ChatGPT did" in their questions, degrading "show your work" to "show your prompt"... |
17:37 |
MTDiscord |
<josiah_wi> I just randomly broke my forks CMake 3.5 CI build due to a bug in an internal CMake module that wasn't fixed until CMake 3.9. I've done a workaround with a comment explaining what's going on. |
17:43 |
Zughy[m] |
Reminder that #12764 should be reviewed and merged as soon as possible to allow people to test it thoroughly |
17:43 |
ShadowBot |
https://github.com/minetest/minetest/issues/12764 -- Sound refactor and improvements by Desour |
18:03 |
|
proller joined #minetest-dev |
18:25 |
|
proller joined #minetest-dev |
18:45 |
sfan5 |
rubenwardy: mesh nodes are copied into the final mesh and entirely unoptimized |
18:45 |
sfan5 |
as for node boxes the answer was already discussed |
18:45 |
rubenwardy |
Why were stairs converted to meshes in mtg? |
18:46 |
rubenwardy |
thanks for the info though |
19:00 |
Krock |
meshes save 4 vertices |
19:00 |
Krock |
actually just 2. It's the overlapping part of the adjacent nodeboxes |
19:01 |
MTDiscord |
<luatic> does Minetest not optimize that out currently? |
19:02 |
Desour |
the amount of vertices is usually not the rendering bottleneck anyway, btw |
19:03 |
|
v-rob joined #minetest-dev |
19:03 |
Krock |
IIRC mesh optimizations are only performed on extruded wield items |
19:04 |
MTDiscord |
<luatic> I don't think extruded wield items are optimized at all |
19:04 |
MTDiscord |
<luatic> they used to be, but they currently aren't |
19:04 |
MTDiscord |
<luatic> Desour: I believe this also saves two tris of the overlapping quad face? |
19:05 |
MTDiscord |
<luatic> not that this would matter terribly much |
19:12 |
|
Wuzzy joined #minetest-dev |
19:22 |
|
v-rob joined #minetest-dev |
19:49 |
|
sugarbeet joined #minetest-dev |
19:58 |
Wuzzy |
!seen nrz |
19:58 |
ShadowBot |
Wuzzy: I saw nrz in #minetest-dev 3 weeks, 2 days, 5 hours, 35 minutes, and 56 seconds ago saying "Sfan5 : if you dont care about history then ez" |
19:58 |
Wuzzy |
!seen nrzkt |
19:58 |
ShadowBot |
Wuzzy: I haven't seen nrzkt in #minetest-dev. |
19:58 |
Wuzzy |
!seen nerzhul |
19:58 |
ShadowBot |
Wuzzy: I haven't seen nerzhul in #minetest-dev. |
19:58 |
Wuzzy |
!tell nrz Almost 5 years ago, you've predicted that Discord will still exist today. Well, you were right, congratulations. XD (I still hate Discord XXXXD) |
19:58 |
ShadowBot |
Wuzzy: OK. |
20:53 |
Pexin |
did I do something wrong in the git naming thread? |
20:54 |
Pexin |
s/thread/issue |
21:25 |
|
MTDiscord1 joined #minetest-dev |
21:28 |
|
Evergreen5 joined #minetest-dev |
21:31 |
|
libera-staff joined #minetest-dev |
22:31 |
|
v-rob joined #minetest-dev |
22:32 |
|
panwolfram joined #minetest-dev |
23:49 |
|
diceLibrarian joined #minetest-dev |
23:58 |
Zughy[m] |
Pexin: you've put two names in one comment |