| Time |
Nick |
Message |
| 00:05 |
|
Eragon joined #minetest |
| 00:09 |
|
sparky4 joined #minetest |
| 00:13 |
|
SFENCE joined #minetest |
| 00:40 |
|
diceLibrarian2 joined #minetest |
| 00:46 |
|
SFENCE joined #minetest |
| 00:48 |
|
liceDibrarian joined #minetest |
| 01:20 |
|
SFENCE joined #minetest |
| 01:41 |
ireallyhateirc |
Doors, toilet to scale: https://files.catbox.moe/wpc8lf.png |
| 01:43 |
|
SFENCE joined #minetest |
| 01:44 |
|
dridron joined #minetest |
| 01:52 |
|
Dyo joined #minetest |
| 02:24 |
|
squeaktoy joined #minetest |
| 02:42 |
Soni |
how does luanti handle scraping? |
| 02:49 |
runxiyu |
Soni: what do you mean by scraping |
| 02:58 |
|
SFENCE joined #minetest |
| 03:13 |
Soni |
like |
| 03:13 |
Soni |
SHODAN |
| 03:26 |
|
erle joined #minetest |
| 03:34 |
|
Verticen joined #minetest |
| 03:37 |
runxiyu |
that doesn't make sense to me |
| 03:55 |
cheapie |
If only this was my bank account balance IRL too: https://cheapiesystems.com/media/images/screenshot_20250101_215318.png |
| 04:59 |
|
Markow joined #minetest |
| 05:00 |
|
MTDiscord joined #minetest |
| 05:40 |
|
erle joined #minetest |
| 05:46 |
|
diceLibrarian2 joined #minetest |
| 06:01 |
|
seasharp joined #minetest |
| 07:46 |
|
chilledfrogs joined #minetest |
| 08:10 |
|
PoochInquisitor joined #minetest |
| 08:14 |
|
tarsovbak joined #minetest |
| 08:32 |
|
gregon joined #minetest |
| 08:32 |
|
jaca122 joined #minetest |
| 08:54 |
|
Glaedr joined #minetest |
| 10:10 |
|
jistr joined #minetest |
| 11:17 |
|
Talkless joined #minetest |
| 11:46 |
|
Hanicef joined #minetest |
| 11:51 |
Soni |
do you know what SHODAN is? |
| 12:04 |
MinetestBot |
[git] sfan5 -> minetest/minetest: Disable failing unit test for the time being 0a67e61 https://github.com/minetest/minetest/commit/0a67e6180d22b73140e5efc89ee9eb6f583edd3c (2025-01-02T11:42:19Z) |
| 12:04 |
|
___nick___ joined #minetest |
| 12:08 |
|
___nick___ joined #minetest |
| 12:35 |
|
sys4 joined #minetest |
| 12:48 |
ROllerozxa |
I do know what SHODAN is |
| 12:49 |
ROllerozxa |
I haven't seen it pick up on my luanti server though. admittedly it is on one port number off from the standard but I don't think it supports pinging with the luanti network protocol like the serverlist does |
| 12:49 |
|
ireallyhateirc joined #minetest |
| 12:58 |
|
___nick___ joined #minetest |
| 13:06 |
|
diceLibrarian2 joined #minetest |
| 13:16 |
|
diceLibrarian2 joined #minetest |
| 13:39 |
|
cx384 joined #minetest |
| 13:49 |
|
nickodd joined #minetest |
| 13:55 |
|
diceLibrarian2 joined #minetest |
| 14:01 |
|
diceLibrarian2 joined #minetest |
| 14:01 |
MinetestBot |
[git] wrrrzr -> minetest/minetest: Rename getMapSettingNoiseParams to getNoiseParams a1b8d20 https://github.com/minetest/minetest/commit/a1b8d20f184ccc59d1c102fbc9b090d66c34e62a (2025-01-02T13:59:19Z) |
| 14:01 |
MinetestBot |
[git] cx384 -> minetest/minetest: Main menu server tab search improvements 15f9d2e https://github.com/minetest/minetest/commit/15f9d2eaaea1efccbbff4f14d7269b67bda95885 (2025-01-02T13:59:33Z) |
| 14:01 |
MinetestBot |
[git] cx384 -> minetest/minetest: Add filters tooltip 5ef41ba https://github.com/minetest/minetest/commit/5ef41badf1fbdc9f2ce8b53f8520e79d8f03447b (2025-01-02T13:59:33Z) |
| 14:01 |
MinetestBot |
[git] cx384 -> minetest/minetest: Fixup e56abca https://github.com/minetest/minetest/commit/e56abca0a07f5bfdaaf3c494586ee4829e02ca5a (2025-01-02T13:59:33Z) |
| 14:01 |
MinetestBot |
[git] (4 newer commits not shown) |
| 14:24 |
|
sparky4 joined #minetest |
| 14:26 |
MinetestBot |
[git] cx384 -> minetest/minetest: Main menu server tab search improvements c4d6240 https://github.com/minetest/minetest/commit/c4d624083d3a907d73cbb1eb8f6501c6cf268b4d (2025-01-02T14:22:49Z) |
| 14:32 |
|
jaca122 joined #minetest |
| 15:10 |
|
silverwolf73828 joined #minetest |
| 15:31 |
|
___nick___ joined #minetest |
| 15:42 |
ireallyhateirc |
Hmmm apparently Minetest had (has?) a limit of 3x3 nodes for a mesh, autobox mod says so at least. On the other hand I just made a mesh that's 2x4 nodes |
| 15:43 |
ireallyhateirc |
Is this info still accurate? |
| 15:46 |
rubenwardy |
I believe it impacts selection and collision boxes but not the visual mesh |
| 15:46 |
rubenwardy |
although cross mapblock might be weird |
| 15:46 |
MTDiscord |
<csperson> iirc it was nodeboxes that had that limit, not meshes |
| 15:55 |
ireallyhateirc |
hmmm can confirm, kinda lame but I guess there's a really good performance rationale for this |
| 15:55 |
rubenwardy |
it's common to break bigger structures up into multiple nodes |
| 15:59 |
ireallyhateirc |
there are several problems with that thoguh: extra work with 3D modelling (when done manually) or using the autobox mod which can produce ugly boxes for "non-watertight" 3D models |
| 15:59 |
ireallyhateirc |
also callbacks are on multiple nodes instead of just one |
| 16:01 |
ireallyhateirc |
in the end I may do an entity for doors |
| 16:07 |
ireallyhateirc |
And on top of it unneeded nodes are created |
| 16:15 |
|
sid0 joined #minetest |
| 16:19 |
Soni |
ROllerozxa: maybe SHODAN itself doesn't, but you could make a script that probes IP addresses for luanti servers yeah? |
| 16:21 |
MTDiscord |
<csperson> who the hell is recreating citidel station in luanti |
| 16:21 |
Soni |
anyway just make the network protocol require a 128-bit token and that should prevent SHODAN-like scripts |
| 16:47 |
ireallyhateirc |
And of course I got a bug when running boxgen and I had to report it |
| 16:48 |
ireallyhateirc |
this is my punishment for using a voxel engine to implement a game with realistic-ish scale |
| 17:01 |
|
bodiccea joined #minetest |
| 17:40 |
|
fluxionary joined #minetest |
| 17:47 |
|
Verticen joined #minetest |
| 18:17 |
|
freq joined #minetest |
| 18:18 |
|
freq joined #minetest |
| 18:18 |
|
freq joined #minetest |
| 18:38 |
|
liceDibrarian joined #minetest |
| 18:39 |
ireallyhateirc |
Entities don't have a collision/selection box size limit? |
| 18:45 |
|
Talkless joined #minetest |
| 18:55 |
Krock |
ireallyhateirc: there's surely a limit for entity <-> entity collisions. Node collision data is gathered from add_area_node_boxes, which does respect the bounding box. |
| 18:55 |
Krock |
the larger the entity, the more CPU time your server will need. It's O(n³) |
| 19:02 |
ireallyhateirc |
well but it won't explode if I make a door entity that's 2x4 nodes? |
| 19:11 |
|
Desour joined #minetest |
| 19:11 |
Krock |
ireallyhateirc: I would recommend using a cooler for the server CPU in general. It should be fine. |
| 19:14 |
MTDiscord |
<luatic> Krock: add_area_node_boxes must make assumptions about the other entity's size however |
| 19:14 |
MTDiscord |
<luatic> which means large boxes won't work |
| 19:15 |
MTDiscord |
<luatic> large node selection boxes will work (the server calculates what the largest possible bounding selection boxes may be) but will wreck performance |
| 19:15 |
MTDiscord |
<luatic> large entity selection boxes will not work because there is prefiltering (and we will want to keep that, esp. when we introduce a spatial index) |
| 19:18 |
ireallyhateirc |
what is considered "large"? I'm having a rough time trying to implement doors that are 2x4 nodes |
| 19:20 |
Krock |
@luatic Huh? Oversized nodes are an issue, but the scanned node range is based on the entity collision box. Hence I think that oversized entity cboxes should be okay ...? |
| 19:23 |
ireallyhateirc |
I'm at mercy of boxgen maintainer |
| 19:27 |
MTDiscord |
<luatic> if you only care about entity - node collisions, sure. but if you care about entity - entity, you won't bump into large entities. |
| 19:30 |
ireallyhateirc |
I'd like to have doors that can be opened under an arbitrary angle. And also so players and entities collide correctly with them |
| 19:31 |
ireallyhateirc |
at least with the fully closed and fully open variants |
| 19:53 |
|
Talkless joined #minetest |
| 20:10 |
|
Talkless joined #minetest |
| 20:12 |
|
Desour_ joined #minetest |
| 20:18 |
ireallyhateirc |
@luatic, so if I want "big" entities then I should be prepared to write my own collision code? Also is there any documented value for "big"? |
| 20:20 |
ireallyhateirc |
in the end I could put placeholder barrier node where the door entity is |
| 20:20 |
ireallyhateirc |
but that's still better than having 6 nodes for doors |
| 20:20 |
ireallyhateirc |
or sorry, 8 |
| 20:22 |
ireallyhateirc |
8 nodes times 10-15 types of doors gives 80-120 nodes just for doors (assuming each has unique mesh rather than just a different texture) |
| 20:42 |
Soni |
can't you use one node per door and a handful of filler nodes shared between all doors |
| 20:47 |
ireallyhateirc |
well, I can and that's probably the most sane solution (unfortunately). But I'm kinda abusing the engine with my game idea and that's my prize. |
| 20:49 |
ireallyhateirc |
But I'll also make the nodes change into entities because I need a few horror features: doors shaking (monster behind), doorknob moving furiously (it tries to get inside D:) and the monster blocking the doors with its hairy dark hands |
| 21:39 |
|
Talkless joined #minetest |
| 21:43 |
|
jluc joined #minetest |
| 22:12 |
|
Verticen joined #minetest |
| 23:33 |
|
panwolfram joined #minetest |
