Time |
Nick |
Message |
00:02 |
|
fling joined #minetest |
00:11 |
|
fling joined #minetest |
01:28 |
|
peterz joined #minetest |
03:23 |
|
toluene joined #minetest |
04:00 |
|
MTDiscord joined #minetest |
04:43 |
|
Verticen_ joined #minetest |
05:41 |
|
calcul0n_ joined #minetest |
06:32 |
|
Lesha_Vel joined #minetest |
07:00 |
|
toluene joined #minetest |
07:04 |
|
olliy joined #minetest |
08:16 |
|
___nick___ joined #minetest |
09:57 |
|
appguru joined #minetest |
10:40 |
|
___nick___ joined #minetest |
10:52 |
|
definitelya joined #minetest |
11:36 |
MTDiscord |
<K.W.555> people use this? |
11:39 |
ROllerozxa |
what does it look like? |
11:40 |
ROllerozxa |
anyways yes, people still use IRC, just because you don't doesn't mean it's dead |
11:47 |
comrad |
it never will be obsolete |
11:53 |
Oblomov |
obligatory xkcd https://xkcd.com/1782/ |
11:55 |
celeron55 |
IRC is like the CB radio of the internet. it won't die |
11:59 |
|
jonadab joined #minetest |
12:00 |
muurkha |
the CB radio of the internet is http://www.icb.net/ |
12:01 |
muurkha |
currently 58 users |
12:03 |
celeron55 |
well, maybe technically, but not socially |
12:04 |
celeron55 |
if i've never heard of it, it can't be the CB of the internet |
12:16 |
muurkha |
being virtually unknown and little-used except by old people seems socially exactly like CB radio :) |
12:21 |
|
debiankaios joined #minetest |
12:30 |
ROllerozxa |
Oblomov: the hover text makes it even funnier, "2078: He announces that he's finally making the jump from screen+irssi to tmux+weechat." xD |
12:31 |
Oblomov |
ROllerozxa: hence the title of this post http://wok.oblomov.eu/tecnologia/ready-for-2078/ |
12:37 |
ROllerozxa |
haha perfect, glad I'm not the only one who reads the XKCD hover texts |
12:47 |
|
proller joined #minetest |
12:52 |
|
definitelya_ joined #minetest |
12:55 |
|
erle joined #minetest |
13:00 |
MTDiscord |
<luatic> you're not :) |
13:00 |
MTDiscord |
<luatic> Anyways folks, I've written a powerful debug mod: https://github.com/appgurueu/dbg |
13:33 |
|
Fixer joined #minetest |
14:10 |
|
ronoaldo joined #minetest |
14:59 |
sfan5 |
reading some MCL2 code and |
14:59 |
sfan5 |
-- TODO: Change meta:get/set_string() to get/set_float() for "last_gametime". |
14:59 |
sfan5 |
-- In Windows *_float() works OK but under Linux it returns rounded unusable values like 449540.000000000 |
14:59 |
sfan5 |
wut |
14:59 |
MTDiscord |
<SX> I don't think restricting /lua chat command to singleplayer is really useful at all and generally limits use. On the other hand I guess ppl running mod with insecure env wont update it anyway befor reading through stuff. |
15:03 |
MTDiscord |
<SX> For example it is extremely rare for me to test in single player mode instead of running dev server in local network, but main thing is that ppl who allow insecure env should undestand risks. Ofc nice to have warning in docs or something but still I don't think generic debugger mod really needs restrictions like that in name of security... |
15:24 |
MTDiscord |
<luatic> Ideally you won't be executing Lua code in the chat anyways, but rather executing /dbg to execute code in the console |
15:26 |
MTDiscord |
<luatic> sfan5: that's particularly odd because gametime always is an integer |
15:27 |
MTDiscord |
<luatic> that said, [gs]et_float should probably use doubles under the hood |
15:30 |
MTDiscord |
<luatic> SX: if you don't want the restriction, simply edit the sources. I had originally made a secure.* setting but MT won't let me register that in settingtypes; an insecure setting would be pointless since that could be changed using /set. |
15:38 |
|
appguru joined #minetest |
15:48 |
sfan5 |
@luatic I assume by rounded they meant the ".00000" fractional part |
15:48 |
sfan5 |
which makes no sense since 123.00000 == 123 |
16:00 |
|
wallabra_ joined #minetest |
16:12 |
|
Verticen_ joined #minetest |
16:26 |
MTDiscord |
<SX> Yes, I kinda meant that editing sources is kind of okayish because ppl who run mods in insecure env without isolated dev server should for sure also read code before installing or updating it. It just seems to me useless to do single player check on mod like that, imo better would be to educate users through configuration for example instead of just making obscure claim about single possible attack vector. |
16:30 |
|
moth__ joined #minetest |
16:32 |
MTDiscord |
<SX> It is kinda making it feel somehow safe to use in servers because dangerous stuff is automatically disabled even while it has no protection agains other mods (which imo is good thing but making it feel like somehow safe isn't) |
16:34 |
MTDiscord |
<luatic> The "Security" section in the Readme covers this. |
16:35 |
MTDiscord |
<luatic> BTW I've just found hecks make the very same claim in #/10206 |
16:35 |
MTDiscord |
<luatic> #10206 |
16:35 |
ShadowBot |
https://github.com/minetest/minetest/issues/10206 -- Add network encryption into minetest |
16:40 |
|
___nick___ joined #minetest |
16:42 |
|
specing joined #minetest |
16:42 |
|
___nick___ joined #minetest |
17:12 |
|
greeter joined #minetest |
17:44 |
|
specing joined #minetest |
17:55 |
|
CowboyLv joined #minetest |
17:56 |
CowboyLv |
Hi |
17:57 |
MTDiscord |
<ROllerozxa> hello |
17:57 |
CowboyLv |
Oh it works. I'm new and I want to know something. |
17:58 |
CowboyLv |
Can you upgrade derived assets from CC3 licence its CC4 variant? |
17:58 |
CowboyLv |
to it's* |
18:00 |
CowboyLv |
Maybe it's a dev-channel only question, but this chat seems empty, so.. idk. My appologies if this is annoying. |
18:03 |
sfan5 |
the dev channel isn't less "empty" and it's also not an easy question |
18:11 |
|
orwell96_m joined #minetest |
18:19 |
* Pexin |
just want write soft ware. *but lawyers watching...!* |
18:19 |
* Pexin |
peers around nervously |
18:23 |
CowboyLv |
@Pexin :DD |
18:27 |
CowboyLv |
Posted an ellaborated question to the dev channel 4 now |
18:36 |
|
rogerm[m] joined #minetest |
18:43 |
|
jadedctrl left #minetest |
18:55 |
|
orwell96_m joined #minetest |
19:02 |
|
orwell96_m joined #minetest |
19:05 |
|
orwell96_m joined #minetest |
19:25 |
|
CowboyLv joined #minetest |
19:33 |
|
specing joined #minetest |
19:44 |
|
specing joined #minetest |
19:47 |
|
Lunatrius joined #minetest |
19:50 |
MTDiscord |
<luatic> CowboyLv: Pretty sure you can't. StackOverflow tried and ultimately gave up. |
19:50 |
MTDiscord |
<luatic> See https://meta.stackexchange.com/questions/333678/was-the-retroactive-change-to-cc-by-sa-4-0-approved-by-stack-exchanges-lawyers |
19:51 |
MTDiscord |
<luatic> > Upgrading from prior versions to 4.0 > > Existing content: > > Who owns the rights? > … > If the contributors, then need permission to relicense. Without permission (via terms of use or otherwise), then that content remains under prior version. |
19:51 |
MTDiscord |
<luatic> => you can't upgrade, you're stuck at 3.0 |
19:52 |
MTDiscord |
<luatic> https://meta.stackexchange.com/questions/344491/an-update-on-creative-commons-licensing |
19:53 |
CowboyLv |
Ok, will either have all assets CC3 or do split licensing originals CC4, derivatives CC3 |
19:55 |
CowboyLv |
One more thing tho. If a sapling has the dirt clump of mtg saplings, and everything else by me, is it still a derivative(remix) or I can have own license there |
20:12 |
|
Verticen_ joined #minetest |
20:13 |
|
orwell96_m joined #minetest |
20:33 |
sfan5 |
you will need a lawyer to answer whether a pixely dirt clump meets the bar of being copyrightable |
20:33 |
sfan5 |
or much easier, just play it safe with licenses |
20:36 |
definitelya_ |
"Uhm, it's actually clay, of the brown variety. ¬_¬" |
20:38 |
muurkha |
potentially not just a lawyer but actually a judge |
20:38 |
MTDiscord |
<Benrob0329> I'm going to start putting "CC-BY 4.0 or later" on my stuff now |
20:38 |
muurkha |
or even an en-banc circuit court rehearing |
20:39 |
MTDiscord |
<Benrob0329> Then again, everything new I make is more or less CC0 |
20:39 |
definitelya_ |
Benrob: Nice |
20:42 |
|
kevinsan joined #minetest |
20:45 |
muurkha |
Benrob0329: that's awesome! |
20:47 |
|
iamweasel joined #minetest |
20:55 |
|
proller joined #minetest |
20:58 |
|
Lunatrius joined #minetest |
21:47 |
|
Reepicheep joined #minetest |
21:50 |
|
Talkless joined #minetest |
21:57 |
|
Quandale joined #minetest |
21:58 |
Quandale |
Hey guys, what mods do you guys recommend installing? |
21:58 |
MTDiscord |
<ROllerozxa> stapled bread is a must |
21:58 |
Quandale |
Okay |
22:02 |
Quandale |
How do I increase the spawn rate of mobs? I have a animal mod installed |
22:04 |
|
Verticen_ joined #minetest |
22:05 |
Quandale |
Hey Verticen_ |
22:13 |
|
est31 joined #minetest |
22:13 |
|
Quandale joined #minetest |
22:35 |
|
panwolfram joined #minetest |
22:36 |
Quandale |
Guys, why are doors taking a while to open on my world? I'm still able to mine and place blocks with no delay, it just seems to be doors opening and closing that delay. |
22:36 |
Quandale |
Also inventory updates |
22:38 |
Quandale |
Update: After a few minutes, delay seems to have gotten much shorter. |
22:38 |
muurkha |
my girlfriend and I have had that problem too, I think there's some kind of desync bug in the netcode |
22:39 |
Quandale |
I'm on a singleplayer world |
22:39 |
muurkha |
it's persisted over several versions |
22:41 |
MTDiscord |
<GreenXenith> Its probably not that dig and place isn't lagging, its likely that its all lagging but client prediction makes dig and place look fine |
22:41 |
Quandale |
huh |
22:42 |
Quandale |
Maybe my computer's just trash |
22:43 |
muurkha |
GreenXenith: yeah, but I dont think it's lag |
22:43 |
muurkha |
I mean it happens even on a single computer, with no chance for packet loss or network retransmissions or latency |
22:54 |
MTDiscord |
<GreenXenith> All I can say is it works on everyone else's machine, which I know isn't helpful but there may still be something about your setup. |
23:01 |
muurkha |
no, we've had the problem on several different machines, on a couple of different LANs, including my old laptop, my new laptop, her desktop, her laptop, and her brother's desktop, and evidently on Quandale's machine too, so it is emphatically not the case that "it works on everyone else's machine" |
23:02 |
muurkha |
It occurs more often when we're playing over the network (rather than on a single host) and more often when the network is not a LAN |
23:03 |
muurkha |
the most relevant symptom is that the log tells us it is dropping messages because there are too many of them, which seems to set up some kind of a metastable desynchronized state |
23:03 |
muurkha |
but we don't know how it arises in the first place |
23:03 |
Quandale |
I'll check the log wherever that is |
23:03 |
MTDiscord |
<GreenXenith> It does, in fact, "work on everyone else's machine", in the sense that the majority of users do not experience this issue |
23:05 |
muurkha |
I don't think we have good enough statistics to know whether that's true, but even if that *were* true, there's an enormous difference between "everyone else" and "the majority of users" |
23:05 |
muurkha |
Quandale: debug.log |
23:06 |
MTDiscord |
<GreenXenith> Such a small difference in semantics does not warrant this level of debate. |
23:06 |
Quandale |
lol, the fanbase is so small that our 7 computers are the majority of users |
23:06 |
muurkha |
oh actually we saw them on the server console, running the server in a terminal. I don't actually remember if we saw them in debug.log |
23:07 |
MTDiscord |
<GreenXenith> Moving right along; it may be useful to find the common link between the machines on which this occurs. My guess is network configuration or client settings. |
23:07 |
MTDiscord |
<Warr1024> Minetest works great on every machine, but only in the sense that it does what you configure it to do. If you configure it to be broken then it faithfully will be. Dropped packets suggest you have something installed that's too heavy for your systems or too poorly written for your network. |
23:08 |
Quandale |
Is it debug.txt? I looked in there and couldn't find "dropped" or "message" or "too many" |
23:08 |
muurkha |
GreenXenith: when you're debugging a problem, you have to be careful not to believe things that aren't true |
23:08 |
muurkha |
Warr1024: I think actually Minetest has bugs from time to time :) |
23:08 |
MTDiscord |
<GreenXenith> Thank you for your estute observations |
23:08 |
muurkha |
Warr1024: but yeah, it could easily be some misbehaved mod that produces lots of network traffic, and then Minetest handles the load shedding badly |
23:09 |
Quandale |
I'll run minetest from the terminal and see what I can find. |
23:09 |
muurkha |
I don't know where to start looking at the network traffic though |
23:09 |
MTDiscord |
<Warr1024> I've found it very easy when modding to introduce storms of packets that make the game seem fine on beefy setups but greatly raise the minimum bar for playability, which bites anyone running systems with any capacity challenges. |
23:10 |
Quandale |
Shoot, I installed my copy of minetest from flathub. It might be a different command that I forgot than just "minetest" |
23:10 |
MTDiscord |
<Warr1024> MT basically blindly trusts mods for a lot of things. If you call set_sky then it will send a "set sky" packet to the client. It doesn't care if you sent the exact same packet 50ms ago, it leaves the modder in charge of making sure that it only calls the API method with a legit change. |
23:11 |
MTDiscord |
<GreenXenith> Flatpak could be half the problem :D |
23:11 |
muurkha |
at one point my best guess was that the total mount of network traffic was actually reasonable but that it was getting buffered up for retransmission over a Zerotier tunnel |
23:11 |
muurkha |
or over Wi-Fi |
23:11 |
muurkha |
and so three seconds of traffic would arrive all at once in a burst |
23:11 |
MTDiscord |
<Warr1024> Flatpak has worked well for me but I've never run it on anemic hardware. Eliminating compiling issues as a variable is one nice thing though. |
23:11 |
muurkha |
but that doesn't explain the cases where it happens on a single machine |
23:11 |
Quandale |
I'll install another copy over apt and see if anything changes |
23:13 |
MTDiscord |
<Warr1024> What is the one common denominator between all the machines you tested but not with other machines that are able to run MT decently? Is it some particular mods or config or something? |
23:14 |
muurkha |
unfortunately I don't have access to any machines that never have this problem |
23:14 |
MTDiscord |
<GreenXenith> It sounded like a few of the machines were also on the same network, though net stuff should be irrelevant on singleplayer |
23:14 |
MTDiscord |
<Warr1024> Is this all about doors taking a while to open? |
23:15 |
muurkha |
doors taking a while to open is the most conspicuous symptom |
23:15 |
Quandale |
Yeah pretty much |
23:15 |
|
Lesha_Vel joined #minetest |
23:15 |
MTDiscord |
<Warr1024> I've noticed slow doors too. I don't even understand that one because I don't think I've seen door delays in NodeCore on single player. |
23:16 |
muurkha |
other symptoms include blocks reappearing after you dig them, damage and breath appearing much later (sometimes 30 seconds or a minute later), etc. |
23:16 |
MTDiscord |
<Warr1024> I assume MTG doors is the common thing here ... maybe MTG doors are just weird. |
23:16 |
Quandale |
I dont really get those symptoms muurkha |
23:16 |
muurkha |
chests take a while to open |
23:16 |
muurkha |
it's possible they might be different problems, Quandale! |
23:16 |
MTDiscord |
<Warr1024> The block reappearing thing happens when server lag is bad enough that prediction gets rolled back. I only see that one when the database is stalling. |
23:17 |
MTDiscord |
<GreenXenith> It also happens on poor network connection |
23:17 |
MTDiscord |
<GreenXenith> Source: my garbage internet |
23:18 |
MTDiscord |
<Warr1024> I had major performance problems with the default SQLite backend on some machines. Enabling journal_mode=wal reduced the problem by like 95%. Migrating to postgresql fixed them wholly. |
23:18 |
muurkha |
oh, interesting! I'll try that when I see the problem again |
23:19 |
MTDiscord |
<GreenXenith> Did we establish if other mods were enabled? |
23:19 |
MTDiscord |
<Warr1024> I noted slow doors when I was testing a map I had converted to MTG from MC. Before that ... well, I hadn't really touched MTG in a few years so I guess I just assumed maybe it had always been like that. |
23:19 |
Quandale |
Oh yeah, other mods were enabled, let me list those off |
23:20 |
MTDiscord |
<GreenXenith> (Vanilla MTG should not have any such issues if on capable hardware) |
23:21 |
Quandale |
armor3d, awards, connected_chests, enderpearl, hangglider, hopper, i3, mg, mobs, mobs_animal, moreblocks, pathv7, skinsdb |
23:21 |
MTDiscord |
<GreenXenith> I ran minetest on a celeron d with 256mb ram the other day for fun, and while it only had about 5fps, there was no noticeable lag |
23:22 |
MTDiscord |
<Warr1024> It just takes one wrong mod to spoil the soup... |
23:22 |
MTDiscord |
<GreenXenith> Mg and pathv7 are both mapgen mods right? |
23:22 |
Quandale |
The issue popped up when mg was installed, maybe it's because there's way too many chests and doors in one chunk |
23:22 |
Quandale |
Also yes both are mapgen |
23:23 |
MTDiscord |
<GreenXenith> Not entirely sure why you'd even want mg over builtin mapgens, as its pretty similar to v6 |
23:24 |
MTDiscord |
<GreenXenith> But regardless, both mods are highly suspect in the case of database issues |
23:24 |
Quandale |
Because it has villages and I wanted villages, I thought the chests would hold uncraftable items at random |
23:24 |
Quandale |
Ender pearls being one of those uncraftable items |
23:24 |
|
erle joined #minetest |
23:25 |
MTDiscord |
<GreenXenith> Ah, so it does. There are standalone village mods, in case you are interested or end up needing to remove mg |
23:25 |
|
z812 joined #minetest |
23:26 |
Quandale |
Okay, I'll install a standalone over mg and see how that goes |
23:27 |
MTDiscord |
<Warr1024> You can generally expect a mod to perform well, as long as your computer is at least as powerful as the one the mod's developer wrote and tested it on. |
23:27 |
MTDiscord |
<Warr1024> For mod soups, it's not exactly like you can just take the maximum but that's not a bad lower bound, at least. |
23:29 |
Quandale |
I installed the people mod on a new world and the same issue is popping up, inventory updates are delayed |
23:29 |
MTDiscord |
<GreenXenith> I recall sokomine mentioning in conversation once how low spec their computer was, so their village mod must perform great :p (joke) |
23:29 |
Quandale |
I'll try a new world with no mods |
23:30 |
muurkha |
villages are cool |
23:30 |
Quandale |
Delay is much less noticable without mods |
23:31 |
MTDiscord |
<GreenXenith> Much less, but still there? How long are we talking? |
23:31 |
Quandale |
Like a single frame |
23:31 |
Quandale |
Conclusion: I'm ditching mapgen mods |
23:32 |
MTDiscord |
<GreenXenith> How delayed was it originally? |
23:32 |
Quandale |
half a second to 2 seconds, outliers were 5 seconds |
23:32 |
MTDiscord |
<GreenXenith> Signing |
23:32 |
MTDiscord |
<GreenXenith> Significant |
23:33 |
MTDiscord |
<GreenXenith> Autocorrect go brrr |
23:34 |
MTDiscord |
<GreenXenith> You shouldn't have to ditch mapgen mods entirely... usually a well made one performs well enough to not cause issues |
23:34 |
MTDiscord |
<GreenXenith> Not only that, mapgen mods shouldn't be causing any lag in already generated areas |
23:34 |
Quandale |
Maybe I'll try my hand at making mods for crafting recipes for uncraftable items. What language are mods usually made in? |
23:35 |
MTDiscord |
<GreenXenith> The game api is lua |
23:35 |
Quandale |
Got it |
23:35 |
MTDiscord |
<Warr1024> Haha, was almost about to say English. |
23:35 |
MTDiscord |
<Warr1024> Actually as I understand it there's no way to translate MT content INTO English so the "no language" language kinda has to be English...? |
23:36 |
MTDiscord |
<Warr1024> I forget. |
23:36 |
MTDiscord |
<Warr1024> Programming an MT mod in anything other than Lua is tricky, and also requires that you write a non-Lua APi bridge at least partially in Lua ... so yeah, best just go with Lua. |
23:37 |
Quandale |
Yep |
23:40 |
Quandale |
I'll try modding another day. For now ima lurk |
23:55 |
muurkha |
you should give it a try! you can probably write your first mod from a modding tutorial in less than an hour |
23:55 |
Quandale |
Okay |
23:55 |
muurkha |
or, you can edit an existing mod you already have installed, for example to add a new crafting recipe |
23:56 |
muurkha |
copy and paste an existing recipe and edit it |
23:56 |
Quandale |
That makes way more sense |
23:56 |
muurkha |
don't worry if you don't have experience with programming, because you can go a long way before you have to write any code more complicated than a function call with some literal data |
23:58 |
Pexin |
Quandale muurkha: how much stuff is in those players' inventories? also including backpacks and nested inventory (digtron crates etc)? |
23:59 |
Quandale |
I spawned in villages so nothing up until the first chest |
23:59 |
Pexin |
serialization of big inventory data bloats the network |