| Time |
Nick |
Message |
| 00:03 |
|
Lesha_Vel joined #minetest |
| 00:20 |
|
erlehmann joined #minetest |
| 00:26 |
|
JerryXiao joined #minetest |
| 01:22 |
Parnikkapore_m |
Guest82: Try Settings → All Settings → Graphics → Menus → Font size (and Monospace font size) |
| 01:22 |
Parnikkapore_m |
Minetest should pick up the display scaling used by your DE tho |
| 01:25 |
|
JerryXiao joined #minetest |
| 01:41 |
|
Oksanaa joined #minetest |
| 01:46 |
|
fluxionary joined #minetest |
| 02:11 |
|
v-rob joined #minetest |
| 02:29 |
ghoti |
So.. When I compile minetest on macos using the instructions provided, everything compiles without difficulty, but when I run the game, I spin uncontrollably. Any idea what's going on? Is this a common thing? |
| 02:33 |
|
grouinos joined #minetest |
| 02:34 |
MTDiscord |
<Jonathon> Yes, see Jordan's instructions |
| 02:34 |
MTDiscord |
<Jonathon> *jordachs |
| 02:39 |
|
Verticen joined #minetest |
| 02:39 |
|
Yad joined #minetest |
| 02:58 |
|
YuGiOhJCJ joined #minetest |
| 03:12 |
ghoti |
Jonathon, I found https://git.minetest.org/Jordach/Minetest and https://github.com/Jordach/minetest but both look like they contain merely old forks. Could you help get me a little closer to my target? |
| 03:15 |
ghoti |
Ah, I have found a useful forum post. Reading... |
| 03:29 |
|
Yad_ joined #minetest |
| 03:38 |
|
v-rob joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 04:14 |
|
Menchers left #minetest |
| 04:56 |
|
lemonzest joined #minetest |
| 05:01 |
|
riff-IRC joined #minetest |
| 05:10 |
Yad |
sfan5: Great to see so many improvements in the changelog for 5.5.0, nice work everybody! Is the IrrlichtMt fork more than just a name change? I thought we were already using a special build of Irrlicht. ;-) |
| 05:52 |
|
v-rob joined #minetest |
| 06:35 |
|
calcul0n joined #minetest |
| 06:50 |
|
unexploredtest[m joined #minetest |
| 07:02 |
|
TomTom joined #minetest |
| 07:42 |
|
Thomas-S joined #minetest |
| 07:42 |
|
Thomas-S joined #minetest |
| 07:44 |
|
Andrew left #minetest |
| 07:46 |
|
v-rob joined #minetest |
| 07:58 |
|
riff-IRC joined #minetest |
| 08:30 |
sfan5 |
Minetest was using Irrlicht 1.8.4 (released 2014?) until now, IrrlichtMt is based on Irrlicht 1.9 (their current development state) and has a whole bunch of stuff thrown out plus a few features/fixes |
| 09:01 |
Oblomov |
it's a bit of a pity seeing a new fork of irrlicht, but I guess the work done by stk on theirs (artica?) wasn't of much use for mt? |
| 09:03 |
celeron55 |
the goal of irrlichtmt is to essentially be completely erased some day. maybe not literally, but that's how it compares to what people usually consider a fork to be |
| 09:04 |
celeron55 |
the only way out from irrlicht is to fork it, and erase and replace it in small bites |
| 09:05 |
celeron55 |
the goal of irrlichtmt is not to develop irrlicht |
| 09:16 |
Oblomov |
celeron55: replace it with what? something entirely custom-tailored for MT? |
| 09:17 |
celeron55 |
nothing, essentially |
| 09:18 |
celeron55 |
adding the necessary things into MT so as to not have any use for irrlicht |
| 09:18 |
celeron55 |
so yes |
| 09:42 |
|
specing_ joined #minetest |
| 09:48 |
Oblomov |
bold-strategy-cotton.jpg |
| 09:53 |
|
GNUHacker joined #minetest |
| 09:59 |
celeron55 |
the stakes aren't that high. i mean, it's irrlicht |
| 10:27 |
Oblomov |
lol 8-) |
| 10:32 |
|
cation21 joined #minetest |
| 10:36 |
erlehmann |
Oblomov i think the biggest mistake in the whole situation is that people see irrlichtmt as they see any other fork of a project, i.e. adding a new direction or new features, but ultimately being the same thing |
| 10:36 |
erlehmann |
this is probably some kind of PR failure lol |
| 10:36 |
Oblomov |
erlehmann: yeah, celeron55 explanation here made things more clear |
| 10:37 |
Oblomov |
I am a bit worried about the transitionary period |
| 10:38 |
erlehmann |
no need to be worried, it's already difficult to figure out the regressions caused by new irrlicht things |
| 10:38 |
erlehmann |
the transitionary period is already here |
| 10:38 |
Oblomov |
yeah, which is why I'm still on 5.4 8-D |
| 10:38 |
erlehmann |
you are not the only one |
| 10:38 |
Oblomov |
the fact that debian hasn't updated is a big factor in that though |
| 10:39 |
erlehmann |
debian is updating some time though |
| 10:39 |
Oblomov |
probably the irrlicht fork will not make it easy for debian to up to 5.5 |
| 10:39 |
kabou |
lol debian stale |
| 10:39 |
erlehmann |
nah, i think they got it? |
| 10:39 |
Oblomov |
erlehmann: I'm on sid and no |
| 10:40 |
kabou |
somebody needs to ping the debian games maintainers |
| 10:40 |
erlehmann |
if you want to make it easier, there are two things you could do: 1. make an irrlichtmt that has a proper history 2. maintain a patchset so that you can compile minetest using irrlicht proper + patches |
| 10:40 |
|
hlqkj joined #minetest |
| 10:40 |
Oblomov |
if the debian maintainers decide to package irrlichtmt separately, it's ging to be a long time before 5.5 gets in, because getting a new package in debian is not easy |
| 10:40 |
erlehmann |
kabou stuff has already happened |
| 10:40 |
kabou |
debian has so many packages, it's entirely understandable that they lag so much |
| 10:41 |
erlehmann |
debian is actually super fast if upstream has their shit together |
| 10:41 |
Oblomov |
especially when it's a fork of something which is already in debian |
| 10:41 |
erlehmann |
i know that from unifont |
| 10:41 |
erlehmann |
debian had it faster than arch linux |
| 10:41 |
erlehmann |
ubuntu, on the other hand, had an old version for half a year |
| 10:41 |
erlehmann |
ubuntu is really bad on updating software, they only do it twice a year lol |
| 10:41 |
kabou |
problem with debian is mostly the gargantuan amount of rules, policies and procedures |
| 10:41 |
Oblomov |
the speed of debian depends largely on who the maintainer is |
| 10:42 |
erlehmann |
kabou FUD |
| 10:42 |
Oblomov |
some stuff are updated instantly at least on experimental, often unstable |
| 10:42 |
erlehmann |
the problem is the lack of people doing things |
| 10:42 |
kabou |
i said that same^^ |
| 10:42 |
kabou |
and lack of people who know all of the hoops to jump through in the right order |
| 10:43 |
kabou |
btw i'm still big fan of debian |
| 10:43 |
kabou |
best distro |
| 10:43 |
Oblomov |
I was wondering why I couldn't up this screenshot on the forum http://labrador.oblomov.eu/nodecore/nodecore-multistack-machine.png then I remembered that I took it it in 4K |
| 10:44 |
erlehmann |
Oblomov what is that? |
| 10:46 |
Oblomov |
erlehmann: nodecore |
| 10:46 |
erlehmann |
Oblomov no, the weird shapes of the nodes |
| 10:47 |
Oblomov |
erlehmann: I'm not sure what exactly you're referring to |
| 10:47 |
Oblomov |
the stacks inside the shelves? |
| 10:47 |
Oblomov |
the shape of the prisms? |
| 10:47 |
erlehmann |
yes the prisms |
| 10:50 |
Oblomov |
those are ... prisms 8-) one of the fundamental building block of machines in nodecore |
| 10:51 |
Oblomov |
erlehmann: you can see an older version of that machine in action here https://youtube.com/playlist?list=PLLoeDdrRGbszibnej4jKQsRCqeg0uya7x |
| 10:52 |
Oblomov |
the weird shape is needed to show which faces are “inputs” and which ones are “outputs” |
| 10:52 |
erlehmann |
i did not know that nodecore had such a thing o.0 |
| 10:53 |
Oblomov |
it's a bit like the mesecons wires |
| 10:54 |
Oblomov |
nodecore is quite fascinating in its design, although it can be quite frustrating in the beginning |
| 10:54 |
Oblomov |
especially since many of the assumptions coming from MC or MTG are invalid |
| 10:55 |
Oblomov |
but when you get to the right tech level, the real fun begins |
| 10:55 |
Oblomov |
(or just hack your way to it with /giveme if you want to go for that 8-D) |
| 10:59 |
|
appguru joined #minetest |
| 11:11 |
|
calcul0n joined #minetest |
| 11:21 |
|
erlehmann joined #minetest |
| 11:23 |
erlehmann |
Oblomov the only non-redstone, non-mesecons automation i have tried so far is the one from inside the box |
| 11:32 |
Oblomov |
haven't played ittb yet |
| 11:34 |
|
appguru joined #minetest |
| 11:36 |
Oblomov |
the nodecore automation mechanism is peculiar because, like most of nodecore, it doesn't offer a wide variety of components, so everything has to be designed from like 3 fundamental blocks (prisms, lenses and hinged panels aka doors) |
| 11:45 |
Oblomov |
I did have to fix an issue in the nodecore game to make sure that these things could be used reliably for things such as detecting when there's new material to process 8-) |
| 12:29 |
|
appguru joined #minetest |
| 12:56 |
|
kamdard__ joined #minetest |
| 13:00 |
|
Fixer joined #minetest |
| 13:08 |
|
Yad joined #minetest |
| 13:09 |
|
Markow joined #minetest |
| 13:16 |
|
proller joined #minetest |
| 13:24 |
|
appguru joined #minetest |
| 13:24 |
CieNTi |
do someone knows, speaking about 'basic_robots' mods, what is the difference between 'normal' or 'coroutine' ? |
| 13:43 |
|
definitelya joined #minetest |
| 13:49 |
|
Yad joined #minetest |
| 13:59 |
erlehmann |
Oblomov do you know about wireworld? https://en.wikipedia.org/wiki/Wireworld |
| 13:59 |
Oblomov |
erlehmann: yeah I read about it in an old sciam |
| 14:00 |
erlehmann |
what is a sciam? |
| 14:00 |
erlehmann |
scientific american? |
| 14:00 |
erlehmann |
Oblomov http://golly.sourceforge.net/ |
| 14:02 |
|
Samson1 joined #minetest |
| 14:02 |
Oblomov |
erlehmann: yes, sciam is scientific american |
| 14:02 |
|
Samson16 joined #minetest |
| 14:07 |
erlehmann |
i think pilzadam may have made a wireworld mod once |
| 14:07 |
erlehmann |
Oblomov https://yewtu.be/watch?v=4Vyp67kOVKs |
| 14:12 |
MTDiscord |
<Warr1024> Wireworld is really cool and was actually one of the candidates I would have considered for NodeCore's logic systems, if (1) it weren't so node-change-intensive, and (2) it hadn't already been done long ago. |
| 14:14 |
erlehmann |
what do you mean with node-change-intensive? |
| 14:14 |
erlehmann |
i mean the prisms also have “electrons” |
| 14:14 |
erlehmann |
they light up |
| 14:15 |
MTDiscord |
<Warr1024> In order to send a steady "1" signal in wireworld, you have to have all affected nodes constantly cycling between empty->head->tail, whereas with nodecore's simple stateful optics, you only need to worry about 0<->1 transitions. |
| 14:15 |
MTDiscord |
<Warr1024> Also, the ability to have a circuit in steady state across block unload boundaries is rather important. |
| 14:18 |
erlehmann |
i see |
| 14:18 |
Samson16 |
rubenwardy |
| 14:18 |
Samson16 |
Okay. |
| 14:18 |
|
Yad joined #minetest |
| 14:20 |
Samson16 |
It says I don't have any permissions. |
| 14:20 |
Samson16 |
Ugh, it's been so long since I was last on Minetest I've practically forgotten everything. |
| 14:24 |
|
Yad joined #minetest |
| 14:26 |
|
peterz joined #minetest |
| 14:27 |
Samson16 |
Is there a way to remove unknown nodes? |
| 14:28 |
MTDiscord |
<ROllerozxa> https://content.minetest.net/packages/AntumDeluge/cleaner/ |
| 14:30 |
rj1 |
i dont care for any of this |
| 14:31 |
rj1 |
i mean yes i do |
| 14:36 |
|
RhineDevil joined #minetest |
| 14:36 |
Samson16 |
Thanks. |
| 14:39 |
Samson16 |
I just got this error message: |
| 14:39 |
Samson16 |
ModError: Failed to load and run script from C:\Users\Samson.COZEN2\Desktop\Portable video games\minetest-5.5.0-win64\minetest-5.5.0-win64\bin\..\mods\plantlife_modpack\youngtrees\init.lua: |
| 14:39 |
Samson16 |
...-win64\bin\..\mods\plantlife_modpack\youngtrees\init.lua:136: attempt to call field 'register_on_generate' (a nil value) |
| 14:39 |
Samson16 |
stack traceback: |
| 14:39 |
Samson16 |
...-win64\bin\..\mods\plantlife_modpack\youngtrees\init.lua:136: in main chunk |
| 14:39 |
Samson16 |
Check debug.txt for details. |
| 14:39 |
|
Samson16 was kicked by ShadowBot: Message flood detected. Use a pastebin like paste.ubuntu.com. |
| 14:40 |
|
Samson16 joined #minetest |
| 14:40 |
Samson16 |
Seriously? Okay. I have an error message. Could someone help me with it? |
| 14:41 |
Samson16 |
Let's see. |
| 14:41 |
Samson16 |
Actually, it's okay. Doesn't matter. |
| 14:46 |
|
Yad joined #minetest |
| 14:53 |
|
Yad joined #minetest |
| 15:13 |
|
erlehmann joined #minetest |
| 15:46 |
|
RhineDevil joined #minetest |
| 16:11 |
|
Talkless joined #minetest |
| 16:26 |
|
v-rob joined #minetest |
| 16:56 |
|
v-rob joined #minetest |
| 17:35 |
|
Yad joined #minetest |
| 17:37 |
|
greeter joined #minetest |
| 17:43 |
|
definitelya joined #minetest |
| 18:03 |
|
appguru joined #minetest |
| 18:03 |
|
Desour joined #minetest |
| 18:10 |
|
v-rob joined #minetest |
| 18:35 |
|
LibreGNU joined #minetest |
| 18:39 |
|
appguru joined #minetest |
| 18:40 |
|
v-rob joined #minetest |
| 18:47 |
|
___nick___ joined #minetest |
| 18:52 |
Pexin |
Samson16 rubenwardy: one time my college sysasmin copied the department auth db, reset a student passwd (who was not present) to check a thing for us (I was a TA), then copied it back afterward, during busy usage hours. heh. |
| 18:54 |
|
grouinos joined #minetest |
| 18:55 |
MTDiscord |
<luatic> Bruh, why'd he "copy" the auth DB? Why not just start a transaction to reset the password? |
| 19:00 |
jonadab |
Not all relational databases support transactions. |
| 19:01 |
jonadab |
Especially if this was a while ago. |
| 19:02 |
Yad |
Does Minetest automatically execute the `init.lua` in every mod on player join? Seems like my code is never being arrived at. |
| 19:03 |
jonadab |
I thought init.lua on enabled mods was executed when the world was loaded, when the server started up. |
| 19:03 |
MTDiscord |
<ROllerozxa> init.lua is indeed executed on world load |
| 19:03 |
Yad |
jonadab: oh right, I meant that, since I'm in singleplayer lol |
| 19:03 |
MTDiscord |
<ROllerozxa> but there is register_on_joinplayer which can run a callback whenever a player joins |
| 19:04 |
Yad |
Roller: Sorry I misspoke |
| 19:04 |
MTDiscord |
<ROllerozxa> oh alright sorry |
| 19:04 |
Yad |
np :) |
| 19:04 |
MTDiscord |
<ROllerozxa> runs away in embarrassment |
| 19:04 |
Yad |
Oh you're only slightly delayed :D |
| 19:04 |
Yad |
the fact remains even a minetest.chat_send_all("Woot") does nothing |
| 19:04 |
MTDiscord |
<luatic> jonadab: Transactions aren't new at all. |
| 19:05 |
Yad |
luatic: Is there some mod list I need to use? |
| 19:05 |
MTDiscord |
<luatic> Yad: Not on player join. You have to use register_on_joinplayer for that. |
| 19:05 |
MTDiscord |
<Jonathon> if your doing it at load time, there is no chat to send it to |
| 19:06 |
MTDiscord |
<luatic> TBH MT should output something in that case |
| 19:06 |
MTDiscord |
<luatic> A warning or something |
| 19:06 |
Yad |
luatic: Yes I meant on world-load. It's singleplayer so those are happening at the same time. |
| 19:06 |
jonadab |
luatic: Depends what you compare them to. But when I was in college, some of the low-end database engines didn't support them. |
| 19:06 |
Yad |
luatic: Oooh, no chat to send to...interesting... |
| 19:06 |
MTDiscord |
<Jonathon> incorrect yad |
| 19:06 |
MTDiscord |
<luatic> Yad: No, those are not happening at the same time. |
| 19:06 |
MTDiscord |
<luatic> Load time is usually "before" join time. |
| 19:07 |
Yad |
luatic: I understand on a technical level yes, two things can't happen at the same time... |
| 19:07 |
Yad |
...but is it not typical to send debug chat tests in the init.lua file of a mod? |
| 19:07 |
sfan5 |
it is not typical because it does not work |
| 19:07 |
Yad |
sfan5: So it seems! :D |
| 19:08 |
MTDiscord |
<Jonathon> if you want to test stuff at load, like stuffing a table full of stuff, just temp make the table global, and use //lua minetest.chat_send_all(dump(global)) to see whats there |
| 19:08 |
jonadab |
There's a way to output messages that will show up in the terminal window where the server is started at the command line, assuming it's started that way (which isn't always the case). |
| 19:08 |
rubenwardy |
you need to use the logger, or delay running until a player joins |
| 19:08 |
jonadab |
And then there's logging. |
| 19:08 |
Yad |
I'm confused because I feel like I've done this before... |
| 19:09 |
rubenwardy |
also, busted is nice for *unit* testing |
| 19:09 |
Yad |
...but I was wrong: it was always inside a minetest.register_on_joinplayer( or similar :D |
| 19:10 |
Yad |
So it's not that I'm using the wrong file `init.lua` I just need to wrap the function call inside another which actually gets called by Minetest. :D |
| 19:14 |
Pexin |
luatic jonadab: it was around 2011 and I don't know the DB detailss |
| 19:35 |
|
proller joined #minetest |
| 20:00 |
|
definitelya_ joined #minetest |
| 20:09 |
|
Talkless joined #minetest |
| 20:13 |
|
v-rob joined #minetest |
| 20:24 |
Yad |
How is it possible that I am getting a "cannot concatenate nil" error when I literally have `if name ~= nil then minetest.register_node(minetest.get_current_modname()..name, {}) end` ? |
| 20:25 |
Yad |
The error indicates it begins on the line with register_node there. |
| 20:26 |
BuckarooBanzai |
Yad: _when_ do you execute that command? On init time or later? ;) |
| 20:26 |
Yad |
BuckarooBanzai: heheheh minetest.register_on_joinplayer(function() |
| 20:27 |
BuckarooBanzai |
uhm, i don't think that's even gonna work :/ |
| 20:27 |
Yad |
Let me see...is there a preferred maximum number of lines I can paste here? |
| 20:27 |
BuckarooBanzai |
pastebin for larger things i guess |
| 20:27 |
Yad |
Maybe a preferred pastebin? |
| 20:27 |
BuckarooBanzai |
or a gist |
| 20:27 |
erlehmann |
Yad, https://mister-muffin.de/paste |
| 20:27 |
Desour |
players join *after* load time, you can't register nodes there |
| 20:28 |
Desour |
and minetest.get_current_modname() is probably nil |
| 20:29 |
Yad |
https://pastebin.com/vzUmQjkC |
| 20:30 |
BuckarooBanzai |
can't you do that without the on_joinplayer callback? looks pretty useless there... |
| 20:30 |
sfan5 |
well then the debug output to chat won't work |
| 20:30 |
sfan5 |
this situations can't be fixed |
| 20:30 |
sfan5 |
-s |
| 20:31 |
Yad |
BuckarooBanzai: the chat_send_all wouldn't do anything ... sfan5 yep |
| 20:31 |
MTDiscord |
<Jonathon> get current modname is nil |
| 20:31 |
MTDiscord |
<Jonathon> and register a node when a player is joining is undefined behavior |
| 20:31 |
Yad |
Jonathon: I added that because of the nil error xD |
| 20:31 |
BuckarooBanzai |
how about `minetest.log`? |
| 20:31 |
Yad |
Jonathon: Ohhh, yes yes yes |
| 20:31 |
Yad |
Can't call register_node in the on_join_player |
| 20:32 |
MTDiscord |
<Jonathon> you can, its just undefined behavior |
| 20:32 |
MTDiscord |
<MisterE> Hmm... wonder what happens |
| 20:32 |
MTDiscord |
<Jonathon> it works, you just have to reload for the node to actually display anything |
| 20:32 |
Yad |
Jonathon: Well...it's nil! :D |
| 20:33 |
MTDiscord |
<Jonathon> no, its not |
| 20:33 |
MTDiscord |
<Jonathon> get current modname is ni; |
| 20:33 |
MTDiscord |
<Jonathon> *nil |
| 20:33 |
MTDiscord |
<Jonathon> read your own error |
| 20:33 |
Yad |
Jonathon: Well that's the error type...I already had it with just "name" and the error said can't concatenate nil |
| 20:33 |
MTDiscord |
<MisterE> What does display anything mean? Is it airlike, 'undefined node', or void like? |
| 20:34 |
MTDiscord |
<MisterE> I mean if you set a node to it |
| 20:34 |
MTDiscord |
<Jonathon> it has the unknown node texture |
| 20:35 |
Yad |
This is the error when I use just `name` without the modname call: https://pastebin.com/rxBjehGB |
| 20:36 |
Yad |
I think you're right that I simply cannot use register_node inside on_join_player |
| 20:37 |
MTDiscord |
<luatic> There are many functions that may only be called at load time. The API should perhaps be more clear about this, but I believe somewhere it says "call these functions only at load time" for the registration funcs |
| 20:38 |
Yad |
luatic: Thanks. By moving the register_on_joinplayer to include only the debug dumps, it works as expected. Complains about the lack of colon rather than mistakenly thinking it's nil |
| 20:38 |
MTDiscord |
<Jonathon> works just fine for me |
| 20:38 |
MTDiscord |
<luatic> minetest.get_current_modname() may only be called at load time as well. You can solve this by storing it in a variable accessible to the functions that need it. Example: local modname = minetest.get_current_modname(); minetest.register_on_joinplayer(function() minetest.chat_send_all(modname .. " loaded!") end). |
| 20:38 |
MTDiscord |
<Jonathon> you need the ":" in front |
| 20:39 |
MTDiscord |
<luatic> Again, please have a close look at the docs. It says "returns the loading mod name" and "when loading" there: https://github.com/minetest/minetest/blob/e54f5e544f27860ba2fa6bfea1a4e1fa3f5d4941/doc/lua_api.txt |
| 20:39 |
MTDiscord |
<Jonathon> again, as i said above, you can get it to work, its just undefined behavior |
| 20:40 |
MTDiscord |
<Jonathon> minetest.register_node(":whateverTheHeckYouWant", {}) works |
| 20:40 |
MTDiscord |
<Jonathon> probably because the engine internally is trying to get modname and failing, but : makes it ignore that |
| 20:43 |
Yad |
luatic: Now I need to know if there's a way to loop on an arbitrary number of dimensions (as opposed to hard-coding for i...for j...for k...etc., where the number of dimensions is known in advance) |
| 20:43 |
MTDiscord |
<luatic> Yes, that's possible. One way to go would be recursion. |
| 20:44 |
MTDiscord |
<luatic> But such algorithms are usually extremely inefficient anyways and thus unsuitable for the tight time bounds imposed by Minetest server steps. |
| 20:45 |
Yad |
luatic: Yes fortunately I would only need to use it for node registration. |
| 20:45 |
MTDiscord |
<luatic> Well, how many loops do you want to nest? |
| 20:45 |
Yad |
luatic: That is precisely what I do not know yet. :) |
| 20:46 |
Yad |
luatic: I'm trying to write my code so that I can add additional variables into my list later. |
| 20:46 |
Yad |
luatic: Without having to re-write the register_node call every time I add a variable. |
| 20:46 |
MTDiscord |
<luatic> There's a max of something like 32^k nodes. Every loop at the very least doubles the amount of nodes I guess. So that's a max of 16 loops, probably more like 15. |
| 20:46 |
Yad |
luatic: Such that the name of each registered node is unique based on its variables. |
| 20:46 |
MTDiscord |
<luatic> But spamming node registrations this way usually isn't a good idea. |
| 20:46 |
|
sobkas joined #minetest |
| 20:47 |
MTDiscord |
<luatic> But why would you need multiple nested for loops to determine unique IDs? Simply increment some counter or the like, or even better, make the name dependent on the variables. |
| 20:47 |
Yad |
luatic: Well I can't imagine needing more than (2^8)*128 = oh...32768 which is exactly what you said xD |
| 20:48 |
Yad |
luatic: I thought it would be lovely to have the nodes actually named according to their attributes. :3 |
| 20:48 |
Yad |
luatic: like flag bits, or a genome |
| 20:49 |
Yad |
luatic: for example, grass_AHFC might have height = 1, seed_count = 8, fibre = 6, heat_resistance = 3 |
| 20:50 |
MTDiscord |
<luatic> Many of these properties sound like they might not need a different node. |
| 20:50 |
Yad |
That is an encouraging thought! |
| 20:50 |
Yad |
But I thought nodes had to be exactly the same as each other? |
| 20:50 |
MTDiscord |
<luatic> Seed count for example: Unless you want a physical indication, you might store this somewhere in the meta. |
| 20:50 |
Yad |
(Until metadata is added ) |
| 20:51 |
Yad |
(lol) |
| 20:51 |
Yad |
I did want physical indications...but maybe that is a bridge too far for Minetest. |
| 20:51 |
MTDiscord |
<MisterE> The only thing you use separate nodes for is different appearances |
| 20:51 |
Yad |
Yeah. |
| 20:51 |
MTDiscord |
<luatic> not quite |
| 20:51 |
MTDiscord |
<luatic> but mostly, yes |
| 20:51 |
MTDiscord |
<luatic> for different colors (or different colors of just parts), you might use color paramtypes |
| 20:52 |
MTDiscord |
<luatic> you might also use separate nodes for other nodedef stuff though |
| 20:52 |
MTDiscord |
<luatic> say different viscosity |
| 20:52 |
Yad |
luatic: and that would be only for colorization, not changing the texture image? |
| 20:52 |
MTDiscord |
<luatic> colorization applies to the texture? |
| 20:52 |
MTDiscord |
<luatic> or parts of the texture |
| 20:52 |
Yad |
I don't want colorization I want different textures xD |
| 20:53 |
MTDiscord |
<MisterE> colori'ation can apply to one (or more) of an over or underlay |
| 20:53 |
Yad |
But you suggested colorization as one of the degrees of freedom in Minetest for the saeme node :) |
| 20:53 |
Yad |
*colorisation ;) |
| 20:53 |
Yad |
*colourisation? |
| 20:53 |
MTDiscord |
<MisterE> @criticalfault42 your area |
| 20:54 |
MTDiscord |
<luatic> Colorization is a degree of freedom |
| 20:54 |
Yad |
It's an interesting idea. I see the general theme is that I should group nodes as much as possible by their non-visual traits in order to have fewer definitions. |
| 20:54 |
Yad |
luatic: Yes, it is. |
| 20:54 |
MTDiscord |
<MisterE> Thank you |
| 20:54 |
MTDiscord |
<luatic> There are more degrees of freedom though |
| 20:54 |
MTDiscord |
<luatic> Say leveled, that might be useful for growing crops |
| 20:55 |
Yad |
luatic: It just doesn't seem very useful to my visual thinking...unless I could use color to represent different plant species or something. |
| 20:55 |
Yad |
Basically I'm simulating the process of developing cultivars. |
| 20:55 |
Yad |
So that "wheat" isn't just already there in the wild. |
| 20:55 |
MTDiscord |
<luatic> Or liquids, or raillike, or all the other drawtypes. All have degrees of freedom usually based on paramtype2. |
| 20:55 |
Yad |
Such that you have to start with grass and selectively breed it into wheat. |
| 20:55 |
MTDiscord |
<luatic> Yad: I was thinking certain properties might be associated with certain colors. |
| 20:56 |
MTDiscord |
<MisterE> Use entities for grnss |
| 20:56 |
MTDiscord |
<MisterE> *grass |
| 20:56 |
MTDiscord |
<luatic> I.e. the more reddish the wheat, the more heat resistant it is or smth |
| 20:56 |
Yad |
MisterE: I was hoping it wouldn't come to that. xD |
| 20:56 |
MTDiscord |
<luatic> Don't use entities |
| 20:56 |
MTDiscord |
<MisterE> Why not? |
| 20:56 |
MTDiscord |
<luatic> Entities are horrible |
| 20:56 |
MTDiscord |
<luatic> Entities are limited |
| 20:56 |
MTDiscord |
<luatic> You may consider using entities for very few grass nodes |
| 20:56 |
Yad |
MisterE: They're highly resource-intensive on the server and the client. |
| 20:56 |
MTDiscord |
<luatic> Perhaps the ones very close to the server, IDK |
| 20:56 |
MTDiscord |
<luatic> Yeah |
| 20:57 |
MTDiscord |
<luatic> Consider a simple 16x16 field of wheat |
| 20:57 |
MTDiscord |
<luatic> Farming often uses large fields |
| 20:57 |
MTDiscord |
<luatic> Not an issue with nodes, very problematic with entities |
| 20:58 |
MTDiscord |
<luatic> This is basically the main issue of Minetest: Entities are very flexible but way too expensive, and nodes are way too "static" but performant |
| 20:59 |
MTDiscord |
<MisterE> hopefully voxel area ents will alleviate that somewhat |
| 20:59 |
MTDiscord |
<luatic> Nah, we need to optimize entities |
| 20:59 |
MTDiscord |
<luatic> in multiple regards |
| 20:59 |
MTDiscord |
<luatic> you already heard of our efforts ;) |
| 20:59 |
MTDiscord |
<MisterE> Yes that too |
| 20:59 |
MTDiscord |
<luatic> VAEs are useful for bulk stuff |
| 21:00 |
MTDiscord |
<MisterE> Yeah I'm really looking forward to them for things like airships |
| 21:11 |
Yad |
luatic: Your remark about nodes vs. entities tradeoff as the central problem with Minetest presently, I think is very apt! :D |
| 21:13 |
Yad |
Are VAEs already a thing? :D I have imagined they must be possible. I don't even know what the letters stand for but it's obvious you're talking about matrices of nodes which matrices can move independently of the main world. |
| 21:15 |
erlehmann |
Yad try this? https://git.minetest.land/Mineclonia/mcl_meshnode |
| 21:15 |
erlehmann |
I believe GreenXenith had something way more advanced |
| 21:18 |
|
v-rob joined #minetest |
| 21:41 |
|
Yad joined #minetest |
| 21:41 |
|
specing_ joined #minetest |
| 21:46 |
MTDiscord |
<luatic> well well, hacks using current entities are already possible |
| 21:46 |
MTDiscord |
<luatic> the cheapest hacks use lots of attached entiteis |
| 21:46 |
MTDiscord |
<luatic> more advanced hacks might generate models and send them as dynamic media |
| 21:46 |
CieNTi |
do someone knows, speaking about 'basic_robots' mods, what is the difference between 'normal' or 'coroutine' ? |
| 21:47 |
MTDiscord |
<luatic> but there's not proper engine support |
| 21:56 |
Desour |
CieNTi: doesn't that mod have help pages in the robot gui? |
| 21:57 |
CieNTi |
yes, but not very verbose, check it -> https://github.com/ac-minetest/basic_robot/blob/master/robogui.lua#L316 |
| 21:57 |
CieNTi |
or maybe is not about the mod and is about Lua ... i'm not new to programming, but noob at all with Lua xD |
| 21:58 |
CieNTi |
"noob at all" .... mmm ... english ... -> im a full noob about lua, is what i meant |
| 21:58 |
Desour |
mh, yes it's not a good help info at all |
| 22:01 |
sfan5 |
well it seems that coroutines can be paused |
| 22:01 |
Desour |
afaik the robots can normally just do one dig operation (and similar) per execution of the robot code. if you do the coroutine thing, it probably uses coroutines to pause the execution of the code and resume it later where it stopped. coroutines can be used for implementing this |
| 22:01 |
sfan5 |
and others can't |
| 22:01 |
MTDiscord |
<luatic> How can I become a member of https://github.com/minetest-mods/ ? |
| 22:01 |
CieNTi |
I see |
| 22:02 |
CieNTi |
but the "pause" ... is like when in an rtos you friendly leave your task to let other task to continue? or it is something like "you have limited ops per cycle, and a pause just wait for the next cycle" ? |
| 22:02 |
CieNTi |
because I cannot find the use of the cooperative point of view |
| 22:03 |
CieNTi |
and I assume in pure Lua, the coroutine is with more than one process in mind ... am I wrong? |
| 22:05 |
sfan5 |
@luatic try to find out if there's a defined process for it |
| 22:05 |
|
Samson1 joined #minetest |
| 22:05 |
Samson1 |
I remember there was a mod which generated a world map view that you could see in a web browser, someone wouldn |
| 22:05 |
Samson1 |
't happen to remember what it's called? |
| 22:06 |
sfan5 |
CieNTi: pausing a Lua coroutine will not block excetion of other code, I assume the robot code works similarly |
| 22:07 |
Desour |
lua's coroutines are just coroutines (<https://en.wikipedia.org/wiki/Coroutine>). IIRC, the robot mod tries to insert some yield operations at loops and co., this is where the preemptiveness comes from |
| 22:07 |
CieNTi |
yeah but that is the point -> "what other code" speaking about the robot itself? I just have a text to be executed, and can be coroutine or not ... but I have no more than one code per robot (books can be used, but still, is a single source of code, and can be coroutine or not) |
| 22:08 |
erlehmann |
Samson1 https://github.com/minetest-mapserver/mapserver |
| 22:08 |
CieNTi |
Desour: checking your link ... |
| 22:08 |
Samson1 |
Thanks. |
| 22:10 |
CieNTi |
Ok read the coroutine wikipedia, and is clear the meaning for me .. but then going out of lua to the robot itself, I still has the same doubt about "what other code" (if so) :S |
| 22:10 |
CieNTi |
thanks Desour |
| 22:10 |
CieNTi |
and sfan5 |
| 22:26 |
Samson1 |
What's the Worldedit command to delete chunks? |
| 22:29 |
sfan5 |
//deleteblocks |
| 22:30 |
sfan5 |
be warned that it may delete or otherwise affect more than the area you have selected |
| 22:31 |
Samson1 |
Even after area pos? |
| 22:33 |
sfan5 |
after what |
| 22:34 |
Samson1 |
Selecting the area I want to delete using //pos1 and //pos2 with World Edit. |
| 22:35 |
sfan5 |
how you select the area does not matter, it can affect more than that |
| 22:35 |
Samson1 |
Huh. |
| 22:35 |
Samson1 |
I used to use WorldEdit to do it and it ONLY cleared the selected area. |
| 22:36 |
Samson1 |
I used to use it all of the time for parts of maps I didn't like. |
| 22:36 |
MTDiscord |
<Jonathon> unless you mean setting to air, then pretty sure its never been that way |
| 22:36 |
Samson1 |
No, not air. |
| 22:37 |
Samson1 |
If I set it to air the map wouldn't regenerate. |
| 22:38 |
Samson1 |
Well, I suppose this will have to do haha! Thanks. :) |
| 22:52 |
Samson1 |
Right, so I've got a weird issue. chunks aren't regenerating after //deleteblocks but also the map appears to have a border now that nothing will generate outside of. |
| 22:53 |
|
Taoki joined #minetest |
| 23:00 |
|
Sven_vB joined #minetest |
| 23:01 |
|
Samson1 joined #minetest |
| 23:13 |
|
Samson1 joined #minetest |
| 23:15 |
Samson1 |
Is there any way I can get the map gen to start generating again? |
| 23:17 |
|
Lesha_Vel joined #minetest |
| 23:18 |
erlehmann |
Samson1 which game are you using? |
| 23:18 |
erlehmann |
Samson1 if you are using mineclone5, it has an “opinionated” map generator idea that could lead to this |
| 23:18 |
erlehmann |
basically, it saves what is generated in mod storage |
| 23:18 |
Samson1 |
No, I'm using V6. |
| 23:19 |
erlehmann |
ah ok |
| 23:20 |
Samson1 |
A tiny corner of the map got corrupted by another mapgen (my fault) along with some other areas, I started deleting and regenerating the chunks until I got to one which crashed Minetest. That was maybe four years ago and the save I'm on now is a much older version from before I started deleting them. |
| 23:25 |
|
kabou joined #minetest |
| 23:28 |
|
kamdard joined #minetest |
| 23:29 |
|
erlehmann joined #minetest |
| 23:32 |
|
panwolfram joined #minetest |
