| Time |
Nick |
Message |
| 00:06 |
|
appguru joined #minetest |
| 00:42 |
|
Eragon joined #minetest |
| 01:16 |
|
cation joined #minetest |
| 01:25 |
|
HumanG33k joined #minetest |
| 03:25 |
|
hephaestus joined #minetest |
| 03:26 |
hephaestus |
hello, I already had a problem with apples that not decaying when you dig all trunks, but I dont remember why? I'm using lumberjack but in another server it not a problem |
| 03:26 |
hephaestus |
oh and I have techpack too |
| 03:36 |
MTDiscord |
<jordan4ibanez> This is quite a condensed problem |
| 03:36 |
MTDiscord |
<jordan4ibanez> A focused one, which mod provided apple trees? |
| 03:49 |
|
liceDibrarian joined #minetest |
| 03:58 |
|
fluxionary joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 05:02 |
|
Leopold joined #minetest |
| 05:12 |
|
nopjmp joined #minetest |
| 05:23 |
|
zleap joined #minetest |
| 05:23 |
|
zleap joined #minetest |
| 05:24 |
|
sparky4 joined #minetest |
| 05:28 |
|
gregon joined #minetest |
| 05:54 |
|
lemonzest joined #minetest |
| 06:09 |
|
TomTom joined #minetest |
| 06:17 |
|
zleap joined #minetest |
| 06:17 |
|
zleap joined #minetest |
| 06:35 |
|
lemonzest joined #minetest |
| 06:39 |
|
tarsovbak joined #minetest |
| 07:23 |
|
Oblomov joined #minetest |
| 07:35 |
|
isAAAc left #minetest |
| 08:14 |
|
mrkubax10 joined #minetest |
| 08:44 |
|
diceLibrarian2 joined #minetest |
| 09:06 |
|
zleap joined #minetest |
| 09:06 |
|
zleap joined #minetest |
| 09:18 |
|
tarsovbak joined #minetest |
| 09:48 |
|
appguru joined #minetest |
| 10:59 |
|
bdju joined #minetest |
| 11:24 |
|
Thelie joined #minetest |
| 11:36 |
|
ireallyhateirc joined #minetest |
| 12:43 |
|
mrkubax10 joined #minetest |
| 12:44 |
|
mrkubax10 joined #minetest |
| 12:49 |
|
Kimapr joined #minetest |
| 12:51 |
|
Kimapr_ joined #minetest |
| 12:54 |
|
Kimapr_ joined #minetest |
| 13:22 |
|
Nusakan joined #minetest |
| 13:33 |
|
PoochInquisitor joined #minetest |
| 13:36 |
|
Kimapr joined #minetest |
| 13:41 |
|
Kimapr joined #minetest |
| 13:43 |
|
gregon joined #minetest |
| 13:44 |
|
tarsovbak joined #minetest |
| 14:14 |
|
bdju joined #minetest |
| 14:17 |
|
seasharp joined #minetest |
| 14:59 |
|
bdju joined #minetest |
| 15:17 |
|
YuGiOhJCJ joined #minetest |
| 15:44 |
ireallyhateirc |
is there something like minetest.register_on_generated but for mapchunks that have already been generated? |
| 15:50 |
ireallyhateirc |
like minetest.on_emerge that would get called every time a chunk finishes being generated and/or is loaded from disk |
| 15:59 |
sfan5 |
there is no generic load block callback, no |
| 16:00 |
|
mrkubax10 joined #minetest |
| 16:01 |
MTDiscord |
<warr1024> You'd have to register an LBM for every node type, and then waste 1023 of the 1024 callbacks you get each mapblock. |
| 16:02 |
sfan5 |
i think you mean 4095/4096 |
| 16:03 |
ireallyhateirc |
what I do currently is running a loop that scans mapchunks around the player and checks if they're loaded |
| 16:04 |
ireallyhateirc |
and then I get a VM for that chunk |
| 16:04 |
ireallyhateirc |
also LBMs are cringe |
| 16:04 |
sfan5 |
what is your use case? |
| 16:05 |
ireallyhateirc |
My mod gets generated mapchunks and generally speaking changes stuff in them using a VM |
| 16:06 |
ireallyhateirc |
one example is changing seasonal soil in Exile |
| 16:06 |
ireallyhateirc |
I also wrote a snow placer using the mod that is fast and works outside of the active area that ABMs operate on |
| 16:07 |
ireallyhateirc |
so: get mapchunks around the player -> get metadata (the chunk has winter soil, snow) -> modify the chunk based on some external trigger (it's summer, we want to remove snow) |
| 16:09 |
MTDiscord |
<warr1024> So you're reinventing ABMs. Cool. |
| 16:09 |
ireallyhateirc |
ABMs were too slow and didn't work well |
| 16:10 |
MTDiscord |
<warr1024> Lol, yeah, ABMs are always slower than reimplementing ABMs until the reimplementation is actually complete. |
| 16:10 |
|
jaca122 joined #minetest |
| 16:10 |
MTDiscord |
<warr1024> The only reliable way I've found to find mapblocks around a player is just to check something like get_node_or_nil for one node inside the mapblock. |
| 16:11 |
MTDiscord |
<warr1024> You also have to make certain assumptions about what mapblocks could be loaded, e.g. that the only ones you care about will form contiguous bubbles around players or something. |
| 16:12 |
ireallyhateirc |
I assign labels to the chunks and store that in mod storage |
| 16:12 |
ireallyhateirc |
so during mapgen I add "has_trees", "has_ocean", etc. |
| 16:12 |
ireallyhateirc |
and then the mod only modifies the chunks of interest |
| 16:13 |
ireallyhateirc |
AFAIK checking a label per chunk is faster than whatever ABMs do |
| 16:14 |
MTDiscord |
<warr1024> Congrats, you reinvented ABM caching too. That kind of thing is not obvious at first. |
| 16:15 |
MTDiscord |
<warr1024> IIRC there's a PR or issue or something to add block-level ABMs (we should call them ANMs just to be consistently wrong); that's what the engine needs to make this sort of thing actually fast at scale... |
| 16:17 |
ireallyhateirc |
still unimpressed |
| 16:18 |
ireallyhateirc |
I wouldn't try to reinvent the wheel if it worked as it should |
| 16:19 |
ireallyhateirc |
ABMs can't be turned off and that's their main con |
| 16:20 |
ireallyhateirc |
my thing works only if it needs to |
| 16:20 |
MTDiscord |
<warr1024> Yeah, it's not easy to impress people when we've been around the track a few times. I've seen a lot of people get excited about a promising approach that's working really well so far, only to find out that the last little bit needed to get to an actually complete working solution is where things suddenly end up worse than the original. |
| 16:21 |
MTDiscord |
<warr1024> Sometimes things can actually be improved upon but there are often reasons why the original is the way it is, and unless you're working on a very narrow specific case, when you try to reinvent an API, you often either run into those, or you make something that's fast but buggy or unreliable. |
| 16:22 |
MTDiscord |
<warr1024> If you do come up with a better way of approaching ABMs than the existing implementation, then we should definitely consider updating or expanding the existing ABM implementation in the engine to support that approach. |
| 16:22 |
Mantar |
warr: check out seasonal changes in Exile's v4 branch sometime |
| 16:23 |
MTDiscord |
<warr1024> what should I be comparing it to? |
| 16:23 |
Mantar |
(we haven't released it to CDB yet as we've got a few things that aren't ready for primetime) |
| 16:23 |
Mantar |
whatever you like, I'm not aware of any MT game that's done seasons on this scale before |
| 16:24 |
Mantar |
oceans freeze out to the horizon, leaves vanish from trees, soils change, plants die |
| 16:25 |
MTDiscord |
<warr1024> Ah, okay, I thought this was a "we did it faster than anybody else" kind of thing. That's kind of tautological if there are no comparisons. |
| 16:26 |
ireallyhateirc |
warr1024, I really tried to use ABMs for that stuff but it was slow and we constantly got warnings about ABMs failing to process mapblocks |
| 16:26 |
ireallyhateirc |
my mods uses VMs to process entire chunks |
| 16:26 |
Mantar |
other mods have done "seasons" but they don't reach our scale because, as ihirc says, abms limit it in several ways |
| 16:27 |
MTDiscord |
<warr1024> What kind of "slow" are we talking about? Like huge lag spikes, or like the season changes taking too long to reach certain mapblocks? |
| 16:27 |
MTDiscord |
<warr1024> It's pretty well known that the ABM scheduler is shit. |
| 16:27 |
ireallyhateirc |
both |
| 16:27 |
Mantar |
yeah |
| 16:27 |
MTDiscord |
<warr1024> A better scheduler could get about 5x the performance on default configuration |
| 16:27 |
Mantar |
lag is a major problem, but so is the short range |
| 16:27 |
MTDiscord |
<warr1024> The problem is that that's just a single fixed one-time speedup. |
| 16:28 |
Mantar |
the shepherd beats abms on both counts |
| 16:28 |
ireallyhateirc |
(my mod that is) |
| 16:28 |
Mantar |
and it still has room for improvement IMO |
| 16:28 |
Mantar |
afk |
| 16:29 |
MTDiscord |
<warr1024> In my case, I've been looking for a general-purpose API to replace ABMs |
| 16:30 |
MTDiscord |
<warr1024> I've found some things, but they all require ABMs too so far 😆 |
| 16:31 |
ireallyhateirc |
my mod is poorly documented for now, so I haven't released it on ContentDB yet |
| 16:32 |
MTDiscord |
<warr1024> Is it exile-specific or does it work for other games? |
| 16:32 |
MTDiscord |
<warr1024> Also, what kind of licensing? |
| 16:32 |
ireallyhateirc |
GPLv3 and is not exile-specific |
| 16:33 |
MTDiscord |
<warr1024> Ah, okay. |
| 16:33 |
ireallyhateirc |
I could consider relicensing if you're interested in my spaghetti code |
| 16:34 |
MTDiscord |
<warr1024> If it actually works for my use-case, and it were MIT, then I might be able to use it. |
| 16:34 |
ireallyhateirc |
well, I can always do dual-licensing |
| 16:35 |
ireallyhateirc |
implementation of seasonal changes using the mod is here: https://codeberg.org/Mantar/Exile/src/branch/v4/mods/nodes_nature/seasons.lua |
| 16:35 |
ireallyhateirc |
and the mod is here: https://codeberg.org/Mantar/Exile/src/branch/v4/mods/mapchunk_shepherd |
| 16:36 |
ireallyhateirc |
Though the code is a huge mess for now and I'm going to clean it up later |
| 16:36 |
MTDiscord |
<warr1024> "mapchunk shepherd" ... does it operate on blocks or chunks? |
| 16:37 |
ireallyhateirc |
chunks |
| 16:37 |
MTDiscord |
<warr1024> oh weird |
| 16:37 |
ireallyhateirc |
would loading blocks in the VM be any faster? |
| 16:38 |
MTDiscord |
<warr1024> I don't know, since I don't know how the mod works |
| 16:38 |
MTDiscord |
<warr1024> usually I check for block-level loading, since there's no guarantee that block loading or unloading actually happens at chunk boundaries |
| 16:39 |
ireallyhateirc |
For simple tasks it can process a chunk in 10ms IIRC |
| 16:39 |
MTDiscord |
<warr1024> as far as I've heard, chunks are only used at mapgen time. |
| 16:40 |
MTDiscord |
<warr1024> I think the last time I messed around with chunk-level stuff was trying to fit the levels in Klots into the fewest chunks I could. Chunk alignment is a bit weird because the 0,0,0 block is actually in the center of the central chunk, not the corner as I had expected. |
| 16:41 |
MTDiscord |
<warr1024> counterintuitive since blocks are aligned the way you'd expect, and the 0,0,0 node IS in the corner of the 0,0,0 block |
| 16:41 |
MTDiscord |
<warr1024> Handling partially-loaded chunks sounds like it could be complicated |
| 16:41 |
ireallyhateirc |
well, another weird thing is that the grid starts at xyz = -32 |
| 16:42 |
ireallyhateirc |
just so Sam can spawn in the middle of the mapchunk rather than in the corner |
| 16:43 |
MTDiscord |
<warr1024> I think that's by design. It seems like they wanted to be able to just generate a single chunk and load all the terrain close to the player. If they were corner-aligned instead of center-aligned then they'd have to generate 8 chunks to be able to show terrain in all direections |
| 16:44 |
ireallyhateirc |
why not for example spawn the player at xyz +32 instead of xyz 0 ? |
| 16:44 |
MTDiscord |
<warr1024> Processing every node in a chunk sounds like a lot of work though. You'd need to defer callbacks into a queue and run a scheduler to avoid them triggering a ton of lag. Either that, or you'd be stuck doing only like contentid->contentid stateless transformations, couldn't access metadata, and couldn't even run very much logic in the callbacks. |
| 16:44 |
ireallyhateirc |
it would probably do the same thing without making grid coordinates weird |
| 16:44 |
ireallyhateirc |
I have a queue and a scheduler |
| 16:45 |
MTDiscord |
<warr1024> tbh they probably just didn't think of it at the time |
| 16:45 |
MTDiscord |
<warr1024> it's not automatically obvious that making grid coordinates weird would have far-reaching consequences |
| 16:47 |
|
mrkubax10 joined #minetest |
| 16:49 |
ireallyhateirc |
sometime this summer I'm going to improve handling of edge cases, possibly add asynch support and write a proper documentation for the mod |
| 16:50 |
MTDiscord |
<warr1024> It looks sort of like it solves a problem that's adjacent to but not overlapping the ones I have in any of my games, but I'll have to see how it progresses. |
| 16:50 |
MTDiscord |
<warr1024> In my case, it's not like there's some kind of global state like season and I have to make a deterministic replacement based on that; it's more like each node responding to its local conditions independently. |
| 16:50 |
ireallyhateirc |
and if you're really interested then I could add a MIT license to that |
| 16:51 |
MTDiscord |
<warr1024> That kind of localism is a big thing in NodeCore in particular. That game has probably like 90 VABMs in the base game or something (probably close to 40 engine ABMs) |
| 16:51 |
ireallyhateirc |
I also use the shepherd for processing moisture movement through soil |
| 16:52 |
MTDiscord |
<warr1024> If you have more than 63 ABMs then the performance suddenly takes a shit, so I forestall that by combining ABMs that have compatible signatures. |
| 16:52 |
|
jaca122 joined #minetest |
| 16:52 |
ireallyhateirc |
the shepherd shares the VM object among multiple workers |
| 16:53 |
ireallyhateirc |
which also improves performance |
| 16:53 |
|
tarsovbak joined #minetest |
| 16:53 |
ireallyhateirc |
anyway, faster ABMs are always welcome |
| 17:00 |
MTDiscord |
<warr1024> ABM performance is ... surprisingly weird. The ABM cache is a huge factor. Including a "neighbors" check is a huge factor. Neighbor inversion can have a rather big impact on performance. |
| 17:02 |
MTDiscord |
<warr1024> I've seen attempts to recreate the API but in pure Lua, which resulted in ... about the same overall performance. I've also attempted "laundering" the callbacks so they aren't all crammed into a single 200ms lagspike, and ... got bad results (spread the bad latency around, but there was still a spike due to the neighbors check, and the overall throughput was lower) |
| 17:08 |
|
Talkless joined #minetest |
| 17:20 |
jonadab |
How parallelizable would the neighbors check be on a multi-core system? |
| 17:30 |
|
mrkubax10 joined #minetest |
| 17:31 |
|
mrkubax10 joined #minetest |
| 17:38 |
MTDiscord |
<luatic> This sounds more like a case for vectorization than for multithreading, since the neighbors check checks a small, constant amount of neighbors. Multithreading ABMs in general sounds like an idea worth trying though; it should be possible to process (at least the insides of) different mapblocks independently. |
| 17:47 |
ireallyhateirc |
btw, why didn't this get any support? https://github.com/minetest/minetest/pull/14258 |
| 17:47 |
ireallyhateirc |
with not so impressive ABM performance I thought it would be good if you could turn off some of them at times when you don't need them |
| 17:48 |
ireallyhateirc |
sfence implemented that thing in response to the issue I started: https://github.com/minetest/minetest/issues/13112 |
| 17:50 |
ireallyhateirc |
We previously used ABMs for our rain soaker that made soil wet during rains in Exile |
| 17:50 |
ireallyhateirc |
but the ABM was always active, even with no rain because there is no way to disable it |
| 17:51 |
ireallyhateirc |
so the ABM ran "if not raining() then return end" for every node? |
| 18:00 |
|
shaft joined #minetest |
| 18:01 |
jonadab |
Not just for soil nodes? |
| 18:04 |
Mantar |
no, he means "every soil node." |
| 18:08 |
ireallyhateirc |
that's about 19200 nodes per mapchunk assuming a 3 nodes tall layer. To prevent lag spikes we increased the interval to around 100s |
| 18:10 |
ireallyhateirc |
with my mod we only run the rain soak worker when it rains and we can do so every 30s without problems |
| 18:15 |
jonadab |
Interesting. |
| 18:15 |
jonadab |
Is rain server-wide or localized? |
| 18:17 |
ireallyhateirc |
server-wide, but I guess I could program it to be localised too |
| 18:17 |
jonadab |
I mean, ABMs only run in loaded-in areas. |
| 18:18 |
jonadab |
So that may not be a big deal, depending on the _number_ of simultaneous players you need to scale to. |
| 18:19 |
ireallyhateirc |
ah, not like that. Currently there's no catch-up because it'd need programming a simulation of rain+evaporation |
| 18:19 |
ireallyhateirc |
but it does work on servers - each player gets rain when it rains and soil becomes wet |
| 18:19 |
ireallyhateirc |
but only land around players gets processed |
| 18:20 |
ireallyhateirc |
but all loaded chunks get processed, not only the "active" area as ABMs |
| 18:22 |
jonadab |
Ah. |
| 18:27 |
shaft |
Most games use player API and the minetest game player model but ship an outdated renamed version. Why? |
| 18:28 |
shaft |
Repixture got updated but still ships the old b3d with the player floating in air during sitting animation, lordofthetest still has the bug where you sometimes don't sit when you just did the right click animation |
| 18:30 |
shaft |
It would be super easy to have working chairs for every single game made with MT engine if not for the horrible mod soup maintainers |
| 18:30 |
shaft |
I just have to let this out somewhere. Sorry. |
| 18:32 |
sfan5 |
<warr1024> The only reliable way I've found to find mapblocks around a player is just to check something like get_node_or_nil for one node inside the mapblock. |
| 18:32 |
sfan5 |
minetest.compare_block_status()? |
| 18:32 |
|
olliy joined #minetest |
| 18:33 |
ireallyhateirc |
yeah I use the what sfan5 just said |
| 18:34 |
sfan5 |
<warr1024> A better scheduler could get about 5x the performance on default configuration |
| 18:34 |
sfan5 |
has someone proposed one? in writing I mean, not code |
| 18:35 |
sfan5 |
ireallyhateirc: reading your usecase it sounds like we should give modders the tools to track active/loaded blocks |
| 18:35 |
sfan5 |
or maybe easier just introduce a global table that mirrors the C++ state? |
| 18:36 |
ireallyhateirc |
whatever works, if not I can still use my hacked together garbage. though would be cool to have something like that |
| 18:36 |
sfan5 |
can you search for an existing issue and if there's one bring it to my attention? or if not, create one? |
| 18:37 |
|
Nusakan joined #minetest |
| 18:38 |
ireallyhateirc |
okay, will do so later |
| 18:41 |
shaft |
Should I just give up on supporting mod soups? |
| 18:44 |
ireallyhateirc |
shaft, yes |
| 18:44 |
shaft |
okay, seems like the sane approach |
| 18:44 |
ireallyhateirc |
generally speaking you shouldn't assume that a single mod can be compatible with multiple games |
| 18:45 |
ireallyhateirc |
just make your mod work for your favourite games |
| 18:45 |
shaft |
It could totally. |
| 18:45 |
ireallyhateirc |
as for player_api, what's the upstream for that? MTG ? |
| 18:45 |
ireallyhateirc |
I just grabbed it for my game from there I believe |
| 18:46 |
shaft |
Yes. MTG is probably the newest while every other game copy pasted some old but renamed version of it with some of them still using the one where the functions are indefault |
| 18:46 |
shaft |
in default |
| 18:46 |
MTDiscord |
<jordan4ibanez> Almost, every other game |
| 18:46 |
ireallyhateirc |
you could ask mod soup maintainers to update their player_api |
| 18:46 |
shaft |
Every game except for node core |
| 18:47 |
shaft |
EVERY SINGLE ONE |
| 18:47 |
ireallyhateirc |
I believe we have it in Exile too, just don't know what version and if it was ever modified |
| 18:47 |
MTDiscord |
<jordan4ibanez> U r very mean I wrote mine from scratch |
| 18:47 |
MTDiscord |
<jordan4ibanez> https://tenor.com/view/la-noire-subtle-lies-lie-lying-gif-23157493 |
| 18:47 |
shaft |
Sure you have. |
| 18:48 |
MTDiscord |
<jordan4ibanez> Yes, because it's written in typescript with only procedural animation off a static model |
| 18:48 |
Mantar |
Exile's was ported from somewhere way back and has now been extensively rewritten |
| 18:48 |
Mantar |
it's misleadingly named at this point, and will probably get folded into an "exile" core mod in the near future |
| 18:49 |
sfan5 |
ireallyhateirc: actually i guess https://github.com/minetest/minetest/issues/14723 does the trick |
| 18:49 |
Mantar |
well, maybe not "extensively" yet, there are still a bunch of inherited funk, but parts of it have been redone |
| 18:49 |
Mantar |
*there is |
| 18:51 |
shaft |
I'm calling out your bullshit. I just registered furniture for exile and player_api works perfectly fine just like in MTG. Good job, keeping it up to date |
| 18:53 |
ireallyhateirc |
sfan5, yeah that would do probably |
| 18:53 |
ireallyhateirc |
"If nodenames are not defined (empty table or nil) the action should be called once on activation with the first pos in the mapblock and a nil node." - so it would get called only once upon activation? |
| 18:54 |
MTDiscord |
<jordan4ibanez> Shaft who are you talking tooooooo |
| 18:54 |
sfan5 |
don't read too much into that proposal |
| 18:55 |
shaft |
Mantar. He says his player API is totally different but it's the same. |
| 18:55 |
shaft |
(which is a good thing) |
| 18:55 |
Mantar |
heh |
| 18:55 |
MTDiscord |
<jordan4ibanez> Hmm |
| 18:55 |
Mantar |
good to hear we still have compatibility there |
| 18:55 |
MTDiscord |
<jordan4ibanez> Mantar use my trash instead https://github.com/jordan4ibanez/forgotten-lands/blob/main/source/animation_station/init.ts |
| 18:56 |
ireallyhateirc |
sfan5, as long as it works for all mapblocks and not only for active mapblocks then that would indeed work for my use case |
| 18:57 |
Mantar |
jordan: looks pretty nice, might be worth replacing our setup |
| 18:57 |
shaft |
What is your "wood" in Exile and how does crafting work. I can add support for your game. |
| 18:58 |
Mantar |
oh we have soft and hard wood logs, and sticks |
| 18:58 |
MTDiscord |
<jordan4ibanez> https://tenor.com/view/who-knew-never-thought-id-get-this-far-plankton-spongebob-gif-16160045 |
| 18:58 |
Mantar |
and fancier wood objects are oiled as well |
| 18:58 |
ireallyhateirc |
sfan5, so "loaded" too in that case, but AFAIK LBMs work only once for each mapblock (they work for blocks that existed before the LBM started working) |
| 18:59 |
Mantar |
and we use rubenwardy's crafting mod for menu crafting, though there are big changes to it in our dev branch |
| 18:59 |
ireallyhateirc |
if so then it wouldn't work for me |
| 18:59 |
sfan5 |
you can already mark LBMs to run every time |
| 19:00 |
Mantar |
see mods/crafting/readme.md for how to register a recipe |
| 19:00 |
Mantar |
I should write up an exile specific crafting.txt and drop it in docs |
| 19:00 |
ireallyhateirc |
sfan5, riiight forgot |
| 19:00 |
ireallyhateirc |
okay so it would work then |
| 19:03 |
shaft |
I only see logs. Can you not process them further? |
| 19:04 |
MTDiscord |
<jordan4ibanez> Have you tried hitting the top with an axe? |
| 19:04 |
Mantar |
nope, we don't do interim "wood" stages. you can add your furniture to the carpentry bench, and that will imply all the tooling and processing |
| 19:04 |
ireallyhateirc |
we have some planks later though? |
| 19:05 |
ireallyhateirc |
or not |
| 19:05 |
Mantar |
we have wooden floors, that's about as close as we get |
| 19:08 |
MTDiscord |
<jordan4ibanez> What about wooden ceilings |
| 19:09 |
MTDiscord |
<jordan4ibanez> :thonking: but then you get to the issue of wooden walls |
| 19:11 |
Mantar |
we have tiled roofs and floors, and log roofs are popular. we have wattle walls too |
| 19:12 |
MTDiscord |
<jordan4ibanez> Any thatched roofing? |
| 19:12 |
Mantar |
yeah thatch is the easiest roof to set up |
| 19:13 |
Mantar |
wattle-and-thatch is my go-to for a first structure, and is the best portable hut atm |
| 19:13 |
MTDiscord |
<jordan4ibanez> That's amazing |
| 19:13 |
Mantar |
and you need a portable hut, the weather will kill you |
| 19:13 |
Mantar |
overland travel is rough |
| 19:50 |
|
silverwolf73828 joined #minetest |
| 20:03 |
shaft |
do you have woods other than wooden floor boards and logs? |
| 20:04 |
Mantar |
sticks are used for many things, they stand in for beams and planks and things |
| 20:04 |
ireallyhateirc |
wattle is made from sticks |
| 20:05 |
ireallyhateirc |
it's a wicker+grass primitive material |
| 20:45 |
|
illwieckz joined #minetest |
| 20:47 |
shaft |
Okay. Exile support is in. Enjoy https://content.minetest.net/packages/shaft/tables_chairs/ |
| 20:48 |
ireallyhateirc |
t-thanks |
| 20:54 |
Mantar |
nice ^_^ b |
| 21:01 |
shaft |
I should take over player_api and force everyone to use the most recent version or something. |
| 21:03 |
ireallyhateirc |
you could extract it from MTG and put as a separate mod on ContentDB |
| 21:04 |
Mantar |
player_api_plus |
| 21:04 |
Mantar |
the plus is for Please Load Update, Slowpokes |
| 21:04 |
shaft |
Player movement should be overhauled too. You guys remember smart moving from Minecraft? |
| 21:05 |
Mantar |
¯\_(ツ)_/¯ |
| 21:05 |
ireallyhateirc |
you mean something like XaEnvironment? |
| 21:05 |
Mantar |
I found Minecraft kinda boring |
| 21:06 |
ireallyhateirc |
yeah, we wouldn't be here if we liked minecraft lol |
| 21:06 |
ireallyhateirc |
btw what are the features of player_api? |
| 21:06 |
ireallyhateirc |
I just shoved it into my Perfect City game (not yet released) and don't know the implications |
| 21:07 |
shaft |
It's a mod that allows climbing, proper swimming, you fold the hands when it detects you're jumping into water, it has a falling animation, you can crawl and swim through 1 block tall holes etc |
| 21:07 |
ireallyhateirc |
ooh that'd be cool to have |
| 21:07 |
ireallyhateirc |
btw if you want to fork player API then make sure it supports different scales of players |
| 21:08 |
ireallyhateirc |
I'm using a player model that's 3.33 nodes high |
| 21:08 |
shaft |
And you can move hand over hand along steel bars over your head |
| 21:08 |
ireallyhateirc |
to make the scale of stuff more natural (similar to what backroom test did) |
| 21:08 |
shaft |
It was pretty popular for parkour maps |
| 21:09 |
shaft |
I think it had had you jump farther too if you ran more before jumping |
| 21:12 |
|
amfl2 joined #minetest |
| 21:13 |
ireallyhateirc |
I'm going to make a character model with knees and elbows, is this supported by player_api? |
| 21:15 |
shaft |
I don't think player api cares. You need to make the animations after all |
| 21:15 |
ireallyhateirc |
that's good |
| 21:16 |
|
appguru joined #minetest |
| 21:23 |
|
illwieckz joined #minetest |
| 21:25 |
Mantar |
we have some of that in Exile, crawling through 1 block holes and a swimming animation |
| 21:27 |
|
amfl2 joined #minetest |
| 22:24 |
|
Noisytoot joined #minetest |
| 22:36 |
|
panwolfram joined #minetest |
| 22:49 |
|
Noisytoot joined #minetest |
| 23:01 |
|
Noisytoot joined #minetest |
| 23:57 |
|
Verticen joined #minetest |
