| Time |
Nick |
Message |
| 00:10 |
|
mmuller joined #minetest |
| 00:40 |
|
Noisytoot joined #minetest |
| 01:17 |
|
smk joined #minetest |
| 01:21 |
|
Kilroy joined #minetest |
| 01:29 |
fluxionary |
today, i discovered lua coroutines. i love them in python, but i'm unsure about whether i want to use coroutines in mods. i've never seen them in mod code i've contributed to (noting that a search shows they're in techage, techpack, lwcomputers, and modlib). |
| 01:30 |
fluxionary |
they *absolutely* make an iterator i'm working on much saner though, if you understand what you're looking at. |
| 01:36 |
fluxionary |
the application is an attempt to make `minetest.get_objects_in_area` faster |
| 02:02 |
|
luk3yx joined #minetest |
| 02:10 |
muurkha |
in Lua (and Squirrel) they're more flexible than Python generators; they have stacks |
| 02:46 |
|
Lesha_Vel joined #minetest |
| 03:05 |
fluxionary |
i agree, they're more powerful than python generators, but python also has equivalent coroutines |
| 03:06 |
fluxionary |
(coroutines equivalent to what lua has) |
| 03:22 |
|
whosit joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 04:06 |
|
est31 joined #minetest |
| 04:11 |
|
lmisu joined #minetest |
| 04:20 |
muurkha |
well, Python has threads, which you can make equivalent to Lua coroutines by adding a single lock to your program so they can't ever preempt each other |
| 04:21 |
muurkha |
but possibly you are incorrectly thinking async and await in Python is equivalent to Lua coroutines; they aren't and you can't simulate Lua coroutines with them |
| 04:21 |
muurkha |
and Python threads are OS threads, so they're orders of magnitude less efficient than Lua coroutines |
| 04:22 |
lissobone |
Hi. |
| 04:24 |
lissobone |
I've finished the checkpoint mod, but I want to know your opinion: when I use the ret pad and teleport back, the velocity with which the player was moving prior teleportation remains. Would it be wiser to limit the freedom of movement and set player velocity to (0, 0, 0)? |
| 04:25 |
lissobone |
Basically, if you were falling and teleported back, you might take fall damage. |
| 04:25 |
lissobone |
Leaving this opens up some room for tricks and hacks. |
| 04:33 |
lissobone |
I'll just add 2 types of checkpoints: one that keeps velocity and one that doesn't. |
| 04:38 |
Blockhead256[m] |
lissobone: it's probably for the best to set an exact point and 0 velocity (or pre-defined, if you must, though with lag that might be bad) to restore the player to. Otherwise the player risks getting stuck in a loop. |
| 04:44 |
lissobone |
Hmm, alright. |
| 04:45 |
lissobone |
I limited the pad use range to 2.7 so the player would have to stand pretty close to it, thus reducing the chance of randomly falling anywhere. |
| 05:21 |
|
Teckla joined #minetest |
| 05:24 |
|
calcul0n joined #minetest |
| 05:44 |
lissobone |
Players, unfortunately, do not support set_velocity, but I hacked my way to winage and remembered how I studied math at school: x - (x + x) equals -x, and adding -x to x (negative velocity plus regular velocity) yields 0 ({x = 0, y = 0, z = 0}). |
| 05:47 |
lissobone |
Here's how to make use of it: 'local vel = player:get_velocity() player:add_velocity(vector.subtract(vel, vector.add(vel, vel)))' |
| 06:10 |
|
definitelya joined #minetest |
| 06:12 |
|
Cork joined #minetest |
| 06:30 |
|
Evergreen joined #minetest |
| 06:49 |
whosit |
lissobone: that is equivalent to: vector.add(vel, vector.multiply(vel, -1)) |
| 06:50 |
lissobone |
Well, that's also true. I had the multiply way in mind, but instead I wanted to subtract the double of the vector from itself. |
| 06:52 |
whosit |
just saying :) |
| 06:53 |
lissobone |
And your way yields a null vector, I believe: vel + -vel = 0. |
| 06:54 |
whosit |
hm? |
| 06:54 |
whosit |
that wasn't the point? to zero out the speed? |
| 06:56 |
lissobone |
Players don't support set_velocity, only add_velocity. |
| 06:57 |
whosit |
you mean player:add_velocity(vector.multiply(player:get_velocity(), -1)) ? |
| 06:58 |
lissobone |
Yes, that. But the documentation page said something about deprecation, maybe I read it wrongly (?). |
| 07:02 |
whosit |
I found only this: https://github.com/minetest/minetest/blob/1b95998d11fb0d51fbfeedc56de4ca2191ad3f7a/doc/lua_api.txt#L7261 |
| 07:06 |
whosit |
was just saying you can do same thing with less steps, which does not matter in this case, except may be less confusing (at least to me) |
| 07:17 |
|
mrkubax10 joined #minetest |
| 07:36 |
|
nore joined #minetest |
| 08:03 |
|
jvalleroy joined #minetest |
| 08:22 |
|
Warr1024 joined #minetest |
| 08:46 |
|
Warr1024 joined #minetest |
| 09:06 |
|
appguru joined #minetest |
| 09:09 |
lissobone |
I have released the checkpoint mod, but -- oops! -- forgot to add the license notice. |
| 09:15 |
MTDiscord |
<luatic> fluxionary: making iterators more convenient is a typical use case of coroutines, go right ahead :) |
| 09:15 |
MTDiscord |
<luatic> in fact the usage of coroutines in modlib is due to iterator utils |
| 09:22 |
lissobone |
Final question. |
| 09:24 |
lissobone |
When a player punches the checkpoint, it blinks with green for approximately 1/√2 (0.717) seconds, then turns red again. Would it be appropriate to set the checkpoint position metadata only after the green period? This opens up room for some cool tricks. |
| 09:24 |
lissobone |
Why am I even asking? |
| 09:28 |
lissobone |
I only wonder how to change anything in the inventory inside a minetest.after() clause. |
| 09:29 |
sfan5 |
@luatic the patch you provided for spiraling-down seems to incomplete, I don't see a single call to register_mapgen_dofile |
| 09:29 |
sfan5 |
+be |
| 09:30 |
MTDiscord |
<luatic> let me check |
| 09:32 |
sfan5 |
even better when you have the changes just push them to a branch |
| 09:32 |
MTDiscord |
<luatic> ah yeah, I had an uncommitted file that didn't make it into the diff |
| 09:43 |
|
Warr1024 joined #minetest |
| 10:01 |
MTDiscord |
<luatic> sfan5: here you go: https://github.com/spiraling-down/game/tree/mapgen-threads (got sidetracked while compiling https://xkcd.com/303/) |
| 10:01 |
MTDiscord |
<luatic> thank you very much for your work, it makes SD almost playable now that the main thread isn't constantly blocked :) |
| 10:06 |
lissobone |
Almost playable? |
| 10:09 |
sfan5 |
fyi you can import files even from the mapgen env, so the sd_map:stuff hack might not be necessary |
| 10:11 |
MTDiscord |
<luatic> lissobone: there's still quite some optimization to be done even if mapgen doesn't block the main thread anymore |
| 10:52 |
|
YuGiOhJCJ joined #minetest |
| 11:30 |
|
appguru joined #minetest |
| 11:47 |
|
BuckarooBanzai joined #minetest |
| 11:54 |
|
peterz joined #minetest |
| 11:56 |
|
Talkless joined #minetest |
| 12:05 |
sfan5 |
@luatic I can't reproduce the error messages you meant |
| 12:06 |
sfan5 |
plus: I spawned inside a node, this text is overlapping: https://0x0.st/HT1p.png, the window with the story didn't disappear automatically |
| 12:06 |
sfan5 |
and the music is quite loud |
| 12:08 |
sfan5 |
and there's z-fighting on sd_map:deco_pebbles_basalt_1 |
| 12:09 |
sfan5 |
respawning drops me far in the sky |
| 12:09 |
sfan5 |
also z-fighting on snow somehow |
| 12:10 |
sfan5 |
plus it's hard to keep track of the information I get dumped on me in the beginning |
| 12:10 |
sfan5 |
at least a "show item name when hotbar selection changes" thing would be nice |
| 12:28 |
|
kevinsan joined #minetest |
| 12:33 |
|
TheSilentLink joined #minetest |
| 12:45 |
|
TheSilentLink joined #minetest |
| 12:55 |
|
Talkless joined #minetest |
| 13:09 |
|
peterz joined #minetest |
| 13:15 |
|
peterz joined #minetest |
| 13:18 |
|
TheSilentLink joined #minetest |
| 13:22 |
|
mrkubax10 joined #minetest |
| 14:06 |
|
olliy joined #minetest |
| 14:09 |
|
qur joined #minetest |
| 14:47 |
|
proller joined #minetest |
| 15:14 |
|
Thomas-S joined #minetest |
| 15:14 |
|
Thomas-S joined #minetest |
| 15:20 |
MTDiscord |
<luatic> sfan5: It currently doesn't error for me either. This was a while ago, you've probably fixed the underlying cause since then. |
| 15:20 |
sfan5 |
I don't think I have but 🤷 |
| 15:27 |
|
appguru joined #minetest |
| 15:43 |
|
Alnotz joined #minetest |
| 16:05 |
|
uyku joined #minetest |
| 16:11 |
|
lissobone joined #minetest |
| 16:13 |
|
sparky4 joined #minetest |
| 16:15 |
|
Desour joined #minetest |
| 16:16 |
|
fluxionary joined #minetest |
| 17:18 |
|
qur joined #minetest |
| 17:49 |
|
mrkubax10 joined #minetest |
| 17:54 |
|
dabbill joined #minetest |
| 18:32 |
|
Wuzzy joined #minetest |
| 18:33 |
Wuzzy |
do i have access to moveresult to players, too? |
| 18:33 |
Wuzzy |
because I want to look at touching_ground (which tells me if an entity is touching the ground) |
| 18:33 |
|
Talkless joined #minetest |
| 18:34 |
sfan5 |
nope |
| 18:34 |
Wuzzy |
Someone reported a strange bug for my game. Apparently the player Y velocity isn't reliably 0 anymore when the FPS is high, like 165. https://codeberg.org/Wuzzy/Glitch/pulls/35 |
| 18:35 |
Wuzzy |
i suppose looking at the player Y velocity to check if im touching the ground isnt a great idea, right? :D |
| 18:36 |
sfan5 |
that's an engine bug, probably |
| 18:37 |
sfan5 |
similar to this https://github.com/minetest/minetest/issues/10962 |
| 18:40 |
Krock |
the movement is really jerky with 800 FPS according to my tests |
| 18:41 |
Wuzzy |
any tips on how to check "this player is standing on the ground?" |
| 18:42 |
Krock |
pos.Y - 0.1 |
| 18:43 |
Wuzzy |
not precise for me. if the player moves vertically, thats not "standing". seems like i kinda am forced to test vertical vel but it goes weird at high FPS |
| 18:46 |
Krock |
when they're moving vertically they have no solid ground under their feet position, neither if the speed is positive |
| 18:47 |
Krock |
it could be that the collision detection does not exactly set the player at the node's surface, resulting in slightly nonzero downwards velocities. these however should still be <0.1 in absolute value |
| 18:49 |
|
uyku joined #minetest |
| 18:50 |
Wuzzy |
funny you say this because 0.1 is exactly the value suggested to me by the PR author :D |
| 18:51 |
Wuzzy |
it still feels off the vel Y value wiggles around 0 instead of staying exactly 0. |
| 18:52 |
Wuzzy |
but yeah the issue in my game seems to be i set the current threshold to 0.001 instead of 0.1 so maybe it was just too low. |
| 19:00 |
|
uyku joined #minetest |
| 19:11 |
|
zuffnick joined #minetest |
| 19:13 |
zuffnick |
wo gibts die besten mods? |
| 19:13 |
Krock |
bei Marcel D'Avis von 1&1 |
| 19:13 |
zuffnick |
:) |
| 19:14 |
zuffnick |
oder bei obi? |
| 19:14 |
Krock |
#minetest-de ist übrigend wür deutschsprachige Diskussionen. hier nur auf English. |
| 19:22 |
|
Ingar joined #minetest |
| 19:29 |
|
uyku joined #minetest |
| 19:30 |
|
appguru joined #minetest |
| 19:39 |
|
sparky4 joined #minetest |
| 19:41 |
|
Juninho joined #minetest |
| 20:01 |
|
imi joined #minetest |
| 20:17 |
|
dabbill joined #minetest |
| 20:24 |
|
Desour joined #minetest |
| 20:25 |
uyku |
Desour: sup |
| 20:25 |
Desour |
hi |
| 20:26 |
uyku |
Desour: send another msg, am testing something |
| 20:26 |
Desour |
ok |
| 20:26 |
uyku |
Yass, it works |
| 20:27 |
uyku |
Are you playing today? |
| 20:27 |
Desour |
maybe |
| 20:40 |
|
Thermoriax joined #minetest |
| 20:45 |
|
uyku joined #minetest |
| 20:45 |
|
lhofhansl joined #minetest |
| 20:46 |
|
proller joined #minetest |
| 20:52 |
MinetestBot |
[git] numberZero -> minetest/minetest: Cull liquid back face on liquid-glasslike interface (#13594) f1feeb3 https://github.com/minetest/minetest/commit/f1feeb319c0a64ee18fb8b29571d5540c084e12c (2023-06-18T20:52:14Z) |
| 20:54 |
|
Guest26 joined #minetest |
| 20:55 |
|
Guest37 joined #minetest |
| 22:34 |
|
panwolfram joined #minetest |
| 22:37 |
|
natewrench joined #minetest |
| 22:54 |
|
sparky4 joined #minetest |
| 23:09 |
|
proller joined #minetest |
| 23:21 |
|
Guest20 joined #minetest |
| 23:22 |
Guest20 |
is there an sdl2 build of minetest for windows? many mouse issues in the non-sdl2 version of it |
| 23:22 |
Guest20 |
seemed like sdl2 might fix it but i can't find builds |
| 23:25 |
Desour |
Guest20: yes |
| 23:25 |
Desour |
build with -DUSE_SDL2=1 |
| 23:25 |
Guest20 |
no i mean ready builds |
| 23:25 |
Guest20 |
i tried to do it myself but idk how to use all this stuff |
| 23:26 |
Desour |
no |
| 23:26 |
Guest20 |
will it switch to sdl2 in the next release at least? |
| 23:27 |
Desour |
possibly |
| 23:27 |
Guest20 |
heres hoping then |
| 23:27 |
Guest20 |
thanks for the helpe |
| 23:27 |
Guest20 |
*help |
| 23:27 |
Desour |
!next |
| 23:27 |
MinetestBot |
Another satisfied customer. Next! |
