| Time |
Nick |
Message |
| 00:02 |
Gustavo6046 |
please deposit your h's in ##h for safekeeping |
| 00:09 |
|
Lesha_Vel joined #minetest |
| 00:16 |
|
harry-wood joined #minetest |
| 00:21 |
|
grouinos joined #minetest |
| 00:22 |
|
Verticen joined #minetest |
| 00:38 |
|
harry-wood joined #minetest |
| 00:53 |
|
harry-wood joined #minetest |
| 00:58 |
|
Hawker joined #minetest |
| 01:17 |
|
RhineDevil^ joined #minetest |
| 01:21 |
MTDiscord |
<MisterE> :wat: |
| 01:29 |
|
harry-wood joined #minetest |
| 01:42 |
|
harry-wood joined #minetest |
| 01:43 |
|
kamdard joined #minetest |
| 01:48 |
|
Gustavo6046_ joined #minetest |
| 01:50 |
|
harry-wood joined #minetest |
| 01:51 |
|
Gustavo6046 joined #minetest |
| 01:56 |
|
fling joined #minetest |
| 02:00 |
|
Gustavo6046_ joined #minetest |
| 02:00 |
|
Hawk777 joined #minetest |
| 02:02 |
|
Gustavo6046 joined #minetest |
| 02:30 |
|
Gustavo6046 joined #minetest |
| 02:33 |
|
Gustavo6046 joined #minetest |
| 02:45 |
|
harry-wood joined #minetest |
| 03:00 |
|
fling joined #minetest |
| 03:35 |
|
RhineDevil joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 04:05 |
|
fling joined #minetest |
| 04:07 |
|
erle joined #minetest |
| 04:18 |
|
Pexin joined #minetest |
| 04:24 |
|
Unit193 joined #minetest |
| 04:52 |
|
Hawk777 joined #minetest |
| 05:01 |
|
riff-IRC joined #minetest |
| 05:08 |
|
fling joined #minetest |
| 05:39 |
|
calcul0n joined #minetest |
| 06:24 |
|
lemonzest joined #minetest |
| 06:28 |
|
Flabb joined #minetest |
| 06:33 |
|
leo_rockway joined #minetest |
| 06:50 |
|
grouinos joined #minetest |
| 07:11 |
|
independent_ joined #minetest |
| 07:13 |
|
fling joined #minetest |
| 07:16 |
sfan5 |
h |
| 07:45 |
|
___nick___ joined #minetest |
| 08:17 |
|
fling joined #minetest |
| 08:20 |
MTDiscord |
<IhrFussel> erle, here is my 2 cents on the "automation" topic: It of course depends on the type of gameplay the server provides but if you play on one where you gain EXP (like mine) for building, mining, killing mobs and more you absolutely DO NOT want any kind of automation you want your players to level up by putting manual effort into the game |
| 08:22 |
MTDiscord |
<IhrFussel> Anything else would be an unfair advantage over others or in other words cheating |
| 08:24 |
|
fling joined #minetest |
| 08:28 |
|
grouinos joined #minetest |
| 08:30 |
|
x_ joined #minetest |
| 08:30 |
MTDiscord |
<IhrFussel> I already implemented anti-cheat mechanics that at least try to prevent the most basic macros from working (e.g. not counting if the player just repeatedly places/digs nodes at the same position) |
| 08:30 |
|
ilomiglio joined #minetest |
| 08:53 |
|
Cork joined #minetest |
| 09:21 |
|
fling joined #minetest |
| 09:31 |
|
kabou joined #minetest |
| 09:40 |
|
RhineDevil^ joined #minetest |
| 09:40 |
|
Sven_vB_ joined #minetest |
| 09:42 |
|
specing_ joined #minetest |
| 10:20 |
|
GNUHacker joined #minetest |
| 10:23 |
|
fling joined #minetest |
| 10:53 |
|
independent_ joined #minetest |
| 10:56 |
|
kabou joined #minetest |
| 10:59 |
|
definitelya joined #minetest |
| 11:03 |
|
tech_exorcist joined #minetest |
| 11:13 |
|
jvalleroy joined #minetest |
| 11:13 |
|
jvalleroy joined #minetest |
| 11:39 |
|
Fixer joined #minetest |
| 12:13 |
|
qur joined #minetest |
| 12:33 |
|
kamdard joined #minetest |
| 12:34 |
independent_ |
My server takes 30 seconds to start. It would be better if it listened for people requesting to join, held them at the waiting screen, and then began loading when the server is ready. |
| 12:38 |
MTDiscord |
<luatic> It would be better if you optimized your startup times |
| 12:38 |
MTDiscord |
<luatic> Figure out the responsible mods |
| 12:40 |
|
harry-wood joined #minetest |
| 12:47 |
independent_ |
Mostly the ones which log a lot |
| 12:47 |
independent_ |
Eh, dosen't really matter |
| 12:52 |
|
harry-wood joined #minetest |
| 12:58 |
|
grouinos joined #minetest |
| 13:17 |
|
RhineDevil joined #minetest |
| 13:24 |
|
debiankaios joined #minetest |
| 13:25 |
|
proller joined #minetest |
| 13:26 |
MTDiscord |
<IhrFussel> 30 seconds sounds unreal...mine takes 10 max to fully restart |
| 13:26 |
|
qur joined #minetest |
| 13:27 |
MTDiscord |
<Warr1024> I'm sure there are ways to make anything arbitrarily slow. Could be a slow mod, could be an issue with equipment... |
| 13:27 |
MTDiscord |
<Warr1024> I once accidentally registered some recipes in a way that like quadrupled the original startup time... |
| 13:27 |
MTDiscord |
<IhrFussel> If it's related to logging my guess is HDD |
| 13:36 |
|
RhineDevil joined #minetest |
| 13:37 |
|
fling joined #minetest |
| 13:50 |
|
lemonzest joined #minetest |
| 14:02 |
|
RhineDevil joined #minetest |
| 14:16 |
|
fluxionary joined #minetest |
| 14:41 |
|
fling joined #minetest |
| 15:14 |
|
Gustavo6046 joined #minetest |
| 15:16 |
|
Gustavo6046 joined #minetest |
| 15:19 |
|
independent_ joined #minetest |
| 15:31 |
|
grouinos joined #minetest |
| 15:36 |
|
ronoaldo joined #minetest |
| 15:44 |
|
Bombo joined #minetest |
| 15:44 |
|
Bombo joined #minetest |
| 16:02 |
|
Extex joined #minetest |
| 16:24 |
|
leo_rockway joined #minetest |
| 16:27 |
|
ludtwig joined #minetest |
| 16:27 |
ludtwig |
Hey, for those who use minetest on wayland, how do you launch it? Since some weeks I can't, "X_CreateWindow" fails |
| 16:30 |
ludtwig |
(for the record, from the arch repos it still works, git main fails) |
| 16:38 |
|
appguru joined #minetest |
| 16:46 |
|
fling joined #minetest |
| 17:17 |
|
Talkless joined #minetest |
| 17:20 |
|
jason101 joined #minetest |
| 17:20 |
jason101 |
how can i nuke old world and create a new one? i dleeted the directory in /var/games/minetest/worlds/oldworld |
| 17:21 |
jason101 |
restarting server made no difference, all of old world contents came back. running minetest on debian in server mode via systemd |
| 17:22 |
sfan5 |
check the logs, the server says where the world that it uses is |
| 17:23 |
sfan5 |
also just for clarity the order is: stop server, delete world, start server |
| 17:27 |
jason101 |
did so. |
| 17:27 |
jason101 |
everything came back after deleting the directory and recreating a new directory empty wit hthe same name as the world world. |
| 17:29 |
erle |
try a new name and look where a new folder gets created |
| 17:29 |
sfan5 |
what does the "ACTION[Main]: World at [...]" line say? |
| 17:29 |
jason101 |
i should not be the one creating the world? I should let the server do that part? |
| 17:31 |
jason101 |
oh, wow, no matter how i rename the world, it's finding the old one somehow |
| 17:32 |
jason101 |
log shows every renamed world directory was loaded. wtf? |
| 17:32 |
jason101 |
on every server restart and attempt. how is it still finding it? |
| 17:32 |
sfan5 |
the worldname is stored inside world.mt in the world folder |
| 17:33 |
jason101 |
every folder in worlds/ is queried for a world.mt file? |
| 17:33 |
sfan5 |
probably |
| 17:33 |
jason101 |
so how can i backup the old world and generate a fresh clean one for the boys? |
| 17:33 |
lemonzest |
gah debian, was a headache getting it to run as a user sevrvice from ~/.minetest/worlds/ |
| 17:33 |
jason101 |
id like to import a map, but i guess there's a difference between worlds and maps. |
| 17:34 |
sfan5 |
store it somewhere as a zip file |
| 17:34 |
sfan5 |
or move it out of the way, e.g. into ~ |
| 17:34 |
sfan5 |
or edit world.mt to "really" rename it |
| 17:34 |
sfan5 |
or extract the new world with a different name? |
| 17:35 |
jason101 |
what do i edit in world.mt? |
| 17:36 |
sfan5 |
look for the line that begins with world_name = |
| 17:36 |
jason101 |
none. |
| 17:36 |
jason101 |
there's no line |
| 17:36 |
jason101 |
for that definition. |
| 17:37 |
sfan5 |
hm |
| 17:37 |
|
sobkas joined #minetest |
| 17:38 |
jason101 |
honestly that was my last attempt that was a dead end, so i arrived here to ask for help. |
| 17:38 |
* jason101 |
shrugs |
| 17:39 |
jason101 |
is it possible or good practice or a practice at all to have multiple worlds stored in the worlds/ folder? |
| 17:40 |
sfan5 |
absolutely |
| 17:40 |
jason101 |
well, then i started to make something slopppy that's backfiring now. i can't create new worlds and keep old ones. |
| 17:41 |
jason101 |
i was hoping to initate server from commandline pointing to the othe world folder |
| 17:41 |
erle |
jason101 it is praxis to have a lot of worlds in the worlds folder. however, do not keep your backups there. i read a bug report where minetest deleted them on upgrade. |
| 17:43 |
sfan5 |
you can tell minetestserver where to look but no idea how that'd work with the systemd unit (debian provides it) |
| 17:43 |
sfan5 |
if nothing works try moving /var/games/minetest/worlds away entirely and see what happens? |
| 17:45 |
rubenwardy |
I use a custom systemd unit pointing to a bash file |
| 17:46 |
rubenwardy |
which has all the args and a custom error reporting system |
| 17:46 |
sfan5 |
erle: I don't know if you can tell yourself, but telling vague scary stories to users is not helpful at all |
| 17:46 |
erle |
sfan5 i will look for the bug report then |
| 17:46 |
erle |
sfan5 sorry for being too vague |
| 17:48 |
|
fling joined #minetest |
| 17:48 |
erle |
sfan5 jason101 here is it. it is “unconfirmed”, so yes, this is a vague scary story. mea maxima culpa https://github.com/minetest/minetest/issues/12069 |
| 17:49 |
sfan5 |
you'd have a point if this wasn't specific to Android |
| 17:49 |
sfan5 |
(which it is) |
| 17:50 |
erle |
oh, that wasn't at all clear to me when remembering it. i should just have looked it up before mentioning it. |
| 17:50 |
erle |
regardless, backups do belong somewhere else, per another justification |
| 17:50 |
erle |
it's not really a backup if it's the same storage |
| 17:51 |
erle |
again, thank you for callilng me out on it, i should really have looked it up. |
| 17:54 |
|
specing_ joined #minetest |
| 18:22 |
|
hn joined #minetest |
| 18:24 |
hn |
just smol question - how it is with contentdb? if i want to create a full operating system as an addon (modifying files + network usage + etc etc. ) generally safe addon but needs perma-insecure env... :/ how it is ? |
| 18:27 |
sfan5 |
not sure what you mean but mods that require extra permissions can be uploaded to CDB fine, users will just have to grant them to actually us it |
| 18:27 |
sfan5 |
use* |
| 18:30 |
|
grouinos joined #minetest |
| 18:36 |
|
Desour joined #minetest |
| 18:54 |
|
fling joined #minetest |
| 18:54 |
|
debiankaios joined #minetest |
| 19:12 |
hn |
i heard they block adding such addon which are forcing insecure env.. |
| 19:13 |
|
Extex joined #minetest |
| 19:16 |
sfan5 |
hmm well maybe it's written down in the a policy, dunno |
| 19:16 |
hn |
yea but im just asking if i'll get banned then |
| 19:17 |
hn |
cause its not my first blocked addon |
| 19:17 |
|
proller joined #minetest |
| 19:19 |
rubenwardy |
Mods requiring insecure env is perfectly fine, there are several |
| 19:19 |
rubenwardy |
But do expect a more in-depth review due to the security implications |
| 19:23 |
hn |
they'd have moderation. detecting of roles etc.. (anyways - night at me - bye) |
| 19:23 |
hn |
tmorrow at discord |
| 19:28 |
Desour |
It would be interesting to have a tag on cdb for mods that require or can use the insec env. |
| 19:28 |
rubenwardy |
yeah, that was planned at some point |
| 19:41 |
|
Fixer joined #minetest |
| 19:43 |
|
Gustavo6046 joined #minetest |
| 19:58 |
|
fling joined #minetest |
| 19:59 |
|
riff-IRC joined #minetest |
| 20:04 |
MTDiscord |
<luatic> Isn't there a "nontrivial installation" tag? That's 90% insecure envs / requires. |
| 20:05 |
MTDiscord |
<luatic> Anways hn: Your mod can't possibly be fully secure, at least not for server use solely based on the fact that MITM always works. And that's ignoring all the flaws MT possibly has. |
| 20:09 |
Desour |
oh, I had not seen the "Complex installation" flag |
| 20:15 |
sfan5 |
rubenwardy: have you made CDB to automatically test-run mods in a sandbox yet? |
| 20:16 |
rubenwardy |
no, I don't plan to |
| 20:16 |
sfan5 |
>:( |
| 20:16 |
rubenwardy |
requires a CDB CLI to start with |
| 20:17 |
rubenwardy |
I'd rather not run untrusted code like that either, even in a sandbox and a container |
| 20:17 |
rubenwardy |
it could run on a separate VPS and use the REST API |
| 20:19 |
sfan5 |
thats being too paranoid, containers are fine |
| 20:20 |
Desour |
3 out of 8 mods on cdb with the "Complex installation" flag (by 2 modders) use the insecure env, and it's used insecurely in all of them... |
| 20:21 |
rubenwardy |
yeah, adv_chat leaks the ie to the modlib table |
| 20:22 |
Desour |
modlib or whatever else any mod has overridden it with * |
| 20:22 |
rubenwardy |
exactly |
| 20:23 |
rubenwardy |
irc leaks require |
| 20:23 |
rubenwardy |
but that's a limitation in the engine |
| 20:23 |
Desour |
character_anim does basically the same mistake |
| 20:24 |
rubenwardy |
luatic: ^ |
| 20:24 |
Desour |
luatic == LMD? |
| 20:25 |
rubenwardy |
yes |
| 20:25 |
|
tech_exorcist joined #minetest |
| 20:27 |
Desour |
I really wonder/worry about security leaks in my require wrapper: https://github.com/Desour/dslib/blob/e8053f605d183493bd031629eb7ac7ae781e41c5/init.lua#L92 |
| 20:28 |
|
kabou joined #minetest |
| 20:29 |
rubenwardy |
perhaps there should be a `ie.require2` should set the ie on the module |
| 20:29 |
rubenwardy |
*that sets |
| 20:29 |
rubenwardy |
as a function env |
| 20:29 |
rubenwardy |
would solve the need to assign require as a global |
| 20:30 |
Desour |
that's also what my wrapper does |
| 20:30 |
Desour |
an unmodified get_modpath in the insec env would be useful too, btw |
| 20:31 |
sfan5 |
it would be much more useful to have an environment that is not impossible to secure |
| 20:32 |
|
Alias joined #minetest |
| 20:37 |
|
ronoaldo joined #minetest |
| 20:38 |
|
riff_IRC joined #minetest |
| 20:38 |
|
ronoaldo joined #minetest |
| 20:52 |
|
riff_IRC joined #minetest |
| 20:55 |
|
ronoaldo joined #minetest |
| 20:59 |
MTDiscord |
<luatic> ^ |
| 20:59 |
MTDiscord |
<luatic> this |
| 20:59 |
MTDiscord |
<luatic> Desour: I've basically given up on trying to make stuff fully secure |
| 21:00 |
MTDiscord |
<luatic> Because technically, everything including the Lua stdlib could be overridden |
| 21:00 |
MTDiscord |
<luatic> for a select few funcs (ie. request_insecure_env) it'll be detected and blocked |
| 21:00 |
MTDiscord |
<luatic> but what about loadfile, for instance? |
| 21:01 |
MTDiscord |
<luatic> the proper way to go IMO would be each mod obtaining it's own environment |
| 21:01 |
MTDiscord |
<luatic> mod interoperability could either work through a shared environment or a require-reimplementation specific to MT |
| 21:02 |
|
fling joined #minetest |
| 21:03 |
|
riff-IRC joined #minetest |
| 21:03 |
Desour |
For loadfile, I am using IE.loadfile and checking a sha256sum before. |
| 21:04 |
Desour |
IE.loadstring* |
| 21:04 |
MTDiscord |
<luatic> Anyways Desour: My mods don't leak the insecure env |
| 21:04 |
MTDiscord |
<luatic> At least not in the global variable kind of way |
| 21:05 |
MTDiscord |
<luatic> I have to temporarily store them in a mod-scoped var due to my architecture (sometimes I also pass it around using loadfile and chunk args) |
| 21:05 |
MTDiscord |
<luatic> But I always make sure to unset it afterwards, see https://github.com/appgurueu/character_anim/blob/master/importer.lua#L11 f.E. |
| 21:05 |
MTDiscord |
<luatic> So it isn't that easily grabbed |
| 21:06 |
MTDiscord |
<luatic> Although by overriding the right modlib callbacks, it would still be pretty easy |
| 21:07 |
MTDiscord |
<luatic> This leaves me with a nice suggestion: Whenever you're free to topologically reorder a couple mods, always prioritize trusted mods |
| 21:07 |
MTDiscord |
<luatic> This means untrusted mods have to run afterwards, unless they somehow get into the dependency chain of trusted mods |
| 21:07 |
MTDiscord |
<luatic> At which point the trusted mods messed up |
| 21:09 |
Desour |
Loading trusted mods and their depends first sounds like a sound thing to do. |
| 21:09 |
sfan5 |
(still just bandaid) |
| 21:10 |
MTDiscord |
<luatic> sfan5: but very decent bandaid |
| 21:10 |
rubenwardy |
you could keep a lot of compatibility with a per-mod environment with some cleverness |
| 21:10 |
MTDiscord |
<luatic> the environment can only have been tampered with by other trusted mods |
| 21:10 |
rubenwardy |
like, using a metatable to capture assignments to _G[the_modname] |
| 21:10 |
MTDiscord |
<luatic> rubenwardy: agreed, but I don't like the performance implications it probably has |
| 21:10 |
rubenwardy |
to then expose in other mod's envs |
| 21:11 |
MTDiscord |
<luatic> plus, what about rawset / rawget |
| 21:11 |
rubenwardy |
and there's keeping support for overriding minetest.* funcs |
| 21:11 |
rubenwardy |
I don't think exposing globals like that is security sensitive, it's a convenience feature |
| 21:17 |
MTDiscord |
<luatic> I just realized I completely forgot to update Cellestial Game |
| 21:18 |
MTDiscord |
<luatic> It didn't get all the shiny modlib updates |
| 21:37 |
|
riff-IRC joined #minetest |
| 21:50 |
|
troller joined #minetest |
| 22:00 |
|
Sven_vB joined #minetest |
| 22:06 |
|
fling joined #minetest |
| 22:07 |
|
ronoaldo joined #minetest |
| 22:07 |
|
riff-IRC joined #minetest |
| 22:08 |
|
troller joined #minetest |
| 22:09 |
|
ronoaldo joined #minetest |
| 22:09 |
|
ilomiglio joined #minetest |
| 22:13 |
|
riff_IRC joined #minetest |
| 22:15 |
|
harry-wood joined #minetest |
| 22:17 |
|
riff_IRC joined #minetest |
| 22:19 |
|
ronoaldo joined #minetest |
| 22:19 |
|
harry-wood joined #minetest |
| 22:21 |
|
riff_IRC joined #minetest |
| 22:29 |
|
ronoaldo joined #minetest |
| 22:32 |
|
ronoaldo joined #minetest |
| 22:33 |
|
panwolfram joined #minetest |
| 22:52 |
|
riff-IRC joined #minetest |
| 22:58 |
|
riff-IRC joined #minetest |
| 22:58 |
|
proller joined #minetest |
| 23:03 |
|
harry-wood joined #minetest |
| 23:05 |
|
riff_IRC joined #minetest |
| 23:13 |
|
riff-IRC joined #minetest |
| 23:14 |
|
Extex joined #minetest |
| 23:20 |
|
riff_IRC joined #minetest |
| 23:23 |
|
riff_IRC joined #minetest |
| 23:25 |
|
Yad joined #minetest |
| 23:25 |
|
AliasAlreadyTake joined #minetest |
| 23:26 |
|
riff_IRC joined #minetest |
| 23:31 |
|
riff-IRC joined #minetest |
| 23:34 |
|
riff-IRC joined #minetest |
| 23:41 |
|
riff-IRC joined #minetest |
| 23:49 |
|
proller joined #minetest |
| 23:58 |
|
Verticen joined #minetest |
