Time |
Nick |
Message |
00:23 |
|
harry-wood joined #minetest |
00:27 |
|
harry-wood joined #minetest |
00:32 |
|
harry-wood joined #minetest |
00:35 |
|
harry-wood joined #minetest |
00:50 |
|
harry-wood joined #minetest |
01:36 |
|
riff-IRC joined #minetest |
01:52 |
|
harry-wood joined #minetest |
02:05 |
|
Lesha_Vel joined #minetest |
02:07 |
|
Lesha_Vel joined #minetest |
04:00 |
|
MTDiscord joined #minetest |
04:00 |
|
behalebabo joined #minetest |
04:44 |
|
est31 joined #minetest |
04:59 |
erle |
i should probably get a palette with more than 12 colors … https://mister-muffin.de/p/kzDx.png https://git.minetest.land/erlehmann/maps/src/branch/master/init.lua |
05:05 |
erle |
at one point snow was yellow by accident (don't eat yellow snow) https://mister-muffin.de/p/mnSF.png |
05:32 |
|
TomTom joined #minetest |
06:01 |
|
nk_ joined #minetest |
06:01 |
nk_ |
hello |
06:02 |
|
Guest48 joined #minetest |
06:02 |
Guest48 |
hello |
06:09 |
|
nk_ left #minetest |
06:18 |
|
calcul0n joined #minetest |
06:23 |
Pexin |
erle: where the huskies go? |
06:37 |
|
lemonzest joined #minetest |
06:44 |
erle |
Pexin, what do you mean? |
08:03 |
|
harry-wood joined #minetest |
08:11 |
|
YuGiOhJCJ joined #minetest |
09:02 |
|
appguru joined #minetest |
09:24 |
|
___nick___ joined #minetest |
09:45 |
|
harry-wood joined #minetest |
09:52 |
settl3r[m] |
Do "/msg"-sent messages get stored, so that the other player receives it after he gets back online in the server it was sent? |
09:54 |
rubenwardy |
No |
09:54 |
rubenwardy |
There are mods for that |
09:54 |
rubenwardy |
But by default no |
09:54 |
|
orwell96 joined #minetest |
10:00 |
settl3r[m] |
Ok. |
10:00 |
settl3r[m] |
By the way, on some servers i saw that some players had physical mailboxes inside the gameworld, can i use these to write them a message which gets stored for them? |
10:00 |
settl3r[m] |
Or are these purely decorative? |
10:02 |
settl3r[m] |
oh.. do i need paper (from papyrus) and a quill (plus ink?) to be able to write a message? |
10:19 |
|
harry-wood joined #minetest |
10:22 |
calcul0n |
settl3r[m], in mtg you can make writable books from paper but you don't need ink |
10:23 |
calcul0n |
and the mailbox is probably the one from homedecor, it lets people put all kind of stuff for the owner |
10:28 |
|
Wuzzy joined #minetest |
10:29 |
Wuzzy |
What do I need to to for Tutorial be allowed on ContentDB again? Like, which mods *specifically* are offending? |
10:29 |
settl3r[m] |
hm.. so i "only" need to cultivate papyrus to be able to put them a book into their mailbox, with my message? |
10:29 |
Wuzzy |
because the Tutorial "game" is currently delisted :( |
10:34 |
calcul0n |
settl3r[m], yes |
10:35 |
calcul0n |
with farming redo you can also make paper from hemp btw |
10:36 |
|
specing_ joined #minetest |
10:45 |
|
alguien joined #minetest |
11:14 |
settl3r[m] |
ok. |
11:14 |
settl3r[m] |
on another topic, how can i revoke my own privs in singleplayer? i want to create a game, where all cheats are disabled forever. |
11:14 |
|
queria joined #minetest |
11:16 |
settl3r[m] |
or can i create a secondary character for that world, which has only <interact> priv? |
11:17 |
settl3r[m] |
in Minecraft, when switching to hardcore, it was impossible to ever switch the gamemode later, i really loved that mode.. is there something similar possible in Minetest? a certain subgame, maybe? |
11:23 |
Wuzzy |
not that i know of |
11:23 |
|
harry-wood1 joined #minetest |
11:30 |
|
orwell96 joined #minetest |
11:30 |
|
definitelya joined #minetest |
11:30 |
|
proller joined #minetest |
11:34 |
MTDiscord |
<MisterE> settlr, |
11:34 |
MTDiscord |
<MisterE> https://tenor.com/view/well-yes-but-actually-no-well-yes-no-yes-yes-no-gif-13736934 |
11:36 |
MTDiscord |
<MisterE> You can delete the creative mod, override the chat commands that grant you privs, probably remove the 'disable damage' checkbox, but then again, you qan revert those chandes again just as easily as you can make them |
11:37 |
MTDiscord |
<MisterE> *can, *changes |
11:37 |
MTDiscord |
<paradust> there's a really simple trick to reverse hard mode on minecraft |
11:38 |
MTDiscord |
<MisterE> You are about to ruin minecraft for him forever O.o |
11:38 |
MTDiscord |
<paradust> >:) |
12:08 |
|
Sven_vB joined #minetest |
12:12 |
|
debiankaios joined #minetest |
12:16 |
erle |
it's really obvious how well fleck handled the shading for maps when you only use flat colors: https://mister-muffin.de/p/4Vqq.png |
12:33 |
|
orwell96_ joined #minetest |
12:39 |
|
Taoki joined #minetest |
13:06 |
|
proller joined #minetest |
13:08 |
|
queria^clone joined #minetest |
13:20 |
|
Fixer joined #minetest |
13:25 |
|
harry-wood joined #minetest |
13:28 |
Pexin |
settl3r[m]: there is a mail mod where you only need to type /mail to get a send/rcv/read UI, and notifies the recipient on login when there are new msgs |
13:31 |
|
lemonzest joined #minetest |
13:31 |
|
specing joined #minetest |
13:31 |
Pexin |
erle: https://m.youtube.com/watch?v=TLIppgE45wM |
13:53 |
|
___nick___ joined #minetest |
13:55 |
|
___nick___ joined #minetest |
14:38 |
|
queria^clone joined #minetest |
14:40 |
|
ronoaldo joined #minetest |
14:52 |
|
est31 joined #minetest |
14:52 |
|
fling joined #minetest |
14:57 |
|
Toothless joined #minetest |
15:16 |
erle |
it's treasure map time hehe https://mister-muffin.de/p/U_HM.png |
15:41 |
|
ghoti_ joined #minetest |
15:42 |
|
dunj3_ joined #minetest |
15:42 |
|
Thomas-S_ joined #minetest |
15:43 |
|
simon8162 joined #minetest |
15:45 |
|
iamwease1 joined #minetest |
15:47 |
erle |
uh, is it true that people have been using texture modifiers to make signs, like savages? |
15:47 |
erle |
i have read stuff to that effect and it seems … unfortunate |
15:47 |
erle |
i mean i am pretty sure it is true, but anyone knows a mod that works like that? |
15:48 |
MTDiscord |
<luatic> Many, many mods render text to textures using the combine texture modifier |
15:48 |
MTDiscord |
<luatic> Player nametag mods do it |
15:49 |
MTDiscord |
<luatic> Basically all signs mods that display text in-world do it too |
15:53 |
|
izzyb joined #minetest |
15:56 |
|
Mc joined #minetest |
15:56 |
|
nore joined #minetest |
15:57 |
|
Ingar joined #minetest |
15:58 |
erle |
luatic that seems horrible |
15:58 |
|
harry-wood joined #minetest |
15:58 |
erle |
luatic why haven't they done it like you or me to it? |
16:02 |
|
wsor4035 joined #minetest |
16:04 |
|
settl3r[m] joined #minetest |
16:05 |
MTDiscord |
<luatic> 1. Because there was no dynamic media when those mods were written |
16:05 |
MTDiscord |
<luatic> 2. Because it works well and supports older versions |
16:06 |
MTDiscord |
<luatic> 3. Because it's rather simpler (and the texture strings don't even get long i.e. for playernames) |
16:06 |
MTDiscord |
<luatic> simple* |
16:07 |
MTDiscord |
<luatic> In fact this may be more efficient due to the fact that [combine can blit arbitrarily high-res glyphs whereas "our" approach requires painting those as pixels. |
16:19 |
MTDiscord |
<GoodClover> erle: How come the world is rendered at a lower resolution (scaled up) to the HUD in those images? |
16:20 |
MTDiscord |
<GoodClover> is there a dial for that? Can't find anything relevant |
16:21 |
erle |
GoodClover undersampling |
16:21 |
erle |
undersampling = 4 |
16:21 |
erle |
in minetest.conf |
16:22 |
erle |
luatic but there are bug reports “minetest can not handle my 3k texture modifiers” … ._. |
16:22 |
|
Verticen joined #minetest |
16:23 |
MTDiscord |
<luatic> erle: this kind of glyph rendering only requires one texture modifier (combine) though |
16:23 |
MTDiscord |
<luatic> as long as the texmod string is below 16k (short string limit), you're golden |
16:23 |
MTDiscord |
<luatic> (although you might want it to be shorter to not put unnecessary strain on the network) |
16:25 |
MTDiscord |
<GoodClover> thanks :) |
16:26 |
MTDiscord |
<luatic> erle: this really doesn't matter at all in comparison to Minetest not even getting acceleration right |
16:26 |
MTDiscord |
<luatic> i.e. Minetest's physics, as simple as they are, are severely broken ._. |
16:29 |
|
Alias joined #minetest |
17:02 |
alguien |
Does anyone know why the (client lua api) minetest.get_node_or_nil would always return nil, even for positions of loaded nodes? |
17:07 |
MTDiscord |
<Warr1024> maybe because the client mods aren't authorized to see those nodes? |
17:07 |
MTDiscord |
<Warr1024> IIRC client mods can be limited in access to nodes, possibly by range from the player or something...? |
17:09 |
alguien |
MTDiscord, this is my local instance, and I (singleplayer) am right next to the node |
17:12 |
erle |
alguien still check CSM restriction settings |
17:14 |
alguien |
erle, I don't have csm_restriction_noderange, could it be that it defaults to 0? |
17:14 |
erle |
no idea what are the csm bits |
17:15 |
|
Talkless joined #minetest |
17:16 |
sfan5 |
it defaults to zero yes |
17:21 |
alguien |
sfan5, does that mean that by default CSMs are effectively blind? |
17:22 |
sfan5 |
yes |
17:22 |
sfan5 |
this was done at the concern of server owners that they could be used for cheating |
17:24 |
alguien |
hmm |
17:25 |
alguien |
i guess they could allow you to look underground... |
17:25 |
alguien |
(i'm thinking if the radius was the same as say reachable radius) |
17:25 |
MTDiscord |
<ZweiKamel> sure they can let you look underground, but then again, so can a malicious minetest fork |
17:25 |
MTDiscord |
<ZweiKamel> like dragonfire |
17:28 |
alguien |
MTDiscord, you mean it's a client-side restriction?! |
17:28 |
sfan5 |
of course |
17:28 |
alguien |
"of course"? what do you mean of course!? |
17:29 |
alguien |
ay |
17:29 |
alguien |
this makes no sense |
17:29 |
alguien |
the malicious person just overrides it |
17:29 |
MTDiscord |
<GoodClover> say that to the server owners that requested the feature |
17:29 |
MTDiscord |
<ZweiKamel> dragonfire is amazing, i use it for all my online banking transactions |
17:29 |
MTDiscord |
<ZweiKamel> it's also pretty good at getting around anticheat |
17:30 |
alguien |
GoodClover: maybe they wanted a server-side feature |
17:30 |
alguien |
like: don't send clients metadata |
17:30 |
alguien |
not, send clients metadata, but tell them not to use it |
17:30 |
MTDiscord |
<Warr1024> DF is not really MT's concern though. |
17:30 |
alguien |
? |
17:30 |
MTDiscord |
<Warr1024> The CSM restrictions are really there for honest players anyway |
17:31 |
alguien |
Warr: well apparently |
17:31 |
definitelya |
Yeah, only certified malicious people can do that. |
17:31 |
sfan5 |
alguien: as it turns out many people don't do things they aren't meant to if it's at least mildly inconvenient |
17:31 |
MTDiscord |
<Warr1024> i.e. anti-cheat isn't just for people who want to cheat no matter what, it also helps keep honest players honest by reducing temptation. |
17:32 |
alguien |
sfan5, i don't think that works |
17:32 |
alguien |
Warr: "isn't just" -> isn't at all |
17:33 |
MTDiscord |
<Warr1024> Some people make anti-cheat features that are designed to prevent people from purposely cheating |
17:33 |
MTDiscord |
<Warr1024> but it's not easy to do very effectively |
17:33 |
alguien |
sfan5, especially once it's default on most or all servers |
17:33 |
sfan5 |
do you know how easy it is for an experienced person to pick a lock? how many people have you asked why they have a lock on their shed/bike/box/...? |
17:33 |
sfan5 |
and do you have locks on yours? |
17:34 |
alguien |
sfan5, i'm sure whoever can install a CSM can follow instructions to install or patch the client |
17:34 |
alguien |
especially if servers by default don't allow them to be superior to others, which is the wish of their hearts |
17:34 |
alguien |
this is silly |
17:35 |
sfan5 |
you are now saying that it won't stop people who really want to do it while earlier you argued it won't stop anyone |
17:35 |
alguien |
did I? |
17:35 |
sfan5 |
Warr: "isn't just" -> isn't at all |
17:35 |
MTDiscord |
<Warr1024> alguien, I think we've heard these arguments all before |
17:35 |
sfan5 |
hmm where did the <...> go |
17:35 |
alguien |
sfan5, read the rest of that sentence |
17:36 |
alguien |
"Everyone" aren't a concern |
17:36 |
MTDiscord |
<Warr1024> It's like saying that the hardware store shouldn't bother selling door locks because they just sell crowbars anyway. |
17:36 |
alguien |
Warr: well you should've heeded them |
17:36 |
MTDiscord |
<Warr1024> We did heed them. |
17:37 |
sfan5 |
alguien: it looks like we agree then, it doesn't stop dedicated people |
17:37 |
sfan5 |
yet that does not make the feature pointless |
17:37 |
alguien |
sfan5, it doesn't stop malicious people whose whole point is to hack |
17:37 |
sfan5 |
correct |
17:37 |
alguien |
well the feature does more harm than good |
17:38 |
alguien |
i doubt server owners specifically asked you: please make my good-willed users blind |
17:38 |
MTDiscord |
<Warr1024> You're arguing with the wrong people anyway |
17:38 |
MTDiscord |
<Warr1024> it's between you and the server owner |
17:38 |
alguien |
Warr: you're defending it |
17:38 |
MTDiscord |
<ZweiKamel> if people want to hack, they use dragonfire. if you don't want them to hack, then somehow find a way to disallow dragonfire |
17:38 |
MTDiscord |
<Warr1024> Apparently they don't think you should be able to use CSMs for this. |
17:38 |
alguien |
Warr: not when it's the default |
17:38 |
alguien |
Warr: only malicious people |
17:39 |
MTDiscord |
<Warr1024> only malicious people what |
17:39 |
alguien |
Warr: mate we're not talking about a specific server, I'm running the game locally |
17:39 |
sfan5 |
then disable it? |
17:39 |
alguien |
apparently I don't want to let myself run CSMs? |
17:39 |
MTDiscord |
<Warr1024> oh, then you have even MORE no excuse for not being able to configure it. |
17:39 |
MTDiscord |
<ZweiKamel> if you want to cheat on a private game, go ahead |
17:39 |
alguien |
sfan5, it's a silly feature |
17:39 |
MTDiscord |
<ZweiKamel> otherwise don't |
17:39 |
MTDiscord |
<ZweiKamel> csms are useful |
17:39 |
MTDiscord |
<Warr1024> If you think the feature is silly then disable it |
17:39 |
MTDiscord |
<ZweiKamel> this ^ |
17:40 |
sfan5 |
disabling it in singleplayer could be an useful suggestion |
17:40 |
alguien |
Warr: you'll vote for my commit to make the default equal to the player reach range? |
17:40 |
sfan5 |
the player cannot reach through solid nodes |
17:40 |
MTDiscord |
<ZweiKamel> why would you disable it in singleplayer and not multiplayer??? |
17:40 |
sfan5 |
neither can the player reliably identify the exact node it is |
17:41 |
alguien |
ZweiKamel? |
17:42 |
MTDiscord |
<Warr1024> I will not vote for your commit because I don't have a vote anyway. |
17:42 |
alguien |
sfan5, that could be further restricted. But to make the CSM blind by default only for honest players is still silly |
17:42 |
alguien |
(either way) |
17:43 |
MTDiscord |
<Warr1024> A range of 1 or 2 (depending on whether the node the player is in counts and whether rounding is done before/after adding range to player pos) seems like it'd be very hard to argue against, though. |
17:44 |
alguien |
Warr: and it would still be useful, as evidenced by the mod basic_robot which can only read_node in radius 1 too |
17:45 |
alguien |
sorry i'm just angry that i can't do what I wanted, It's not your fault |
17:46 |
alguien |
probably |
17:48 |
alguien |
I'm not sure about sfan5 but I think the rest of you are off the hook XD |
17:48 |
MTDiscord |
<Warr1024> It's open-source software, it's going to have bugs, including issues where nobody can agree whether it's a bug or whether it's right and it'd be a bug the other way around. If you can find a good compromise position that has strong rationale why it's better than the extremes, then there's still a decent chance of effecting change. |
17:49 |
alguien |
Warr: your suggestion of 1 being default is good |
17:49 |
MTDiscord |
<Warr1024> My concern of course include how the rounding is done |
17:50 |
alguien |
Warr I'm sure it can be done right |
17:50 |
|
aeminux joined #minetest |
17:51 |
MTDiscord |
<Warr1024> Can? Definitely. Will? Well ... you increase the chances of that if you do it yourself ? |
17:52 |
MTDiscord |
<Warr1024> Is there a definitive and preferably succinct guide to how to contribute somewhere, walking through the issue -> concept approval -> PR -> review -> 2 approvals lifecycle? |
17:53 |
alguien |
Warr: you'll do that for me? Thanks, you're the best XD |
17:54 |
MTDiscord |
<Warr1024> Well, I guess on some level that IS the guide ... but maybe a bit TOO succinct :-) |
17:57 |
MTDiscord |
<Warr1024> Generally, at least, last time I checked, PRs are supposed to start their lifecycle as an issue and the proposed solution needs to be concept-approved before a PR will be considered, so it's best to do that before you start digging too deep into the code. People may raise some concerns that hadn't been considered and you may want to change the approach. |
17:57 |
alguien |
More sillyness, ceck this out: I join a server, and my csm gets run, and a node returns the info. A few seconds later I guess it's only then that the restriction kicks in and I don't get the info. Ayayayayay puerto rico |
17:57 |
alguien |
thanks for the tip Warr |
17:58 |
definitelya |
c:)> |
17:59 |
alguien |
oh no, it's not that, my bad |
17:59 |
alguien |
it's that i was checking the grass node which was at the exact same position as I was. You know, radius 0? XD |
17:59 |
alguien |
so I can check nodes, i just have to glitch into them hah |
18:00 |
alguien |
yep, just checked a solid one now hehehe |
18:01 |
alguien |
so technically and practically it's not as bad as i thought |
18:02 |
MTDiscord |
<IhrFussel> The point of those CSM restrictions simply is to have SOME type of protection against possible cheaters on servers...of course there will always be those who know how to edit c++ source code or download cheat clients but the majority of users does not |
18:03 |
MTDiscord |
<IhrFussel> There are no cheat clients for Android either I think so those issues only exist on desktop |
18:03 |
erle |
Wuzzy do you have a lua algorithm for bresenhams axe? |
18:03 |
erle |
Wuzzy like, displaying “uses left” in tooltips |
18:04 |
alguien |
IhrFussel: bleh |
18:04 |
alguien |
client side security ¬_¬ |
18:04 |
MTDiscord |
<IhrFussel> Or at least if any of the cheat clients does work on mobile it is not in the Play Store and needs to be downloaded from somewhere else which again almost nobody will do |
18:05 |
alguien |
MTDiscord, fine, maybe a random "PVPer" doesn't get to one-shot people... |
18:06 |
erle |
IMO the point of the CSM restrictions are to calm down server ownners who couldn't distinguish assistive technology from game-breaking cheats ever |
18:07 |
MTDiscord |
<ROllerozxa> what happened to never trusting the client, implement server-side anticheat |
18:07 |
MTDiscord |
<IhrFussel> Many crucial things cannot be cheated from the client no matter how much you modify it like HP, damage, breath, punch interval |
18:07 |
erle |
alguien this gives you a lot of CSMs and more APIs than vanilla minetest, don't cheat https://repo.or.cz/waspsaliva.git |
18:07 |
MinetestBot |
[git] Wuzzy2 -> minetest/minetest: DevTest: Fix broken PNG textures eabf057 https://github.com/minetest/minetest/commit/eabf05758e3ba5f6f4bb1b8d1d1f02179b84e410 (2022-05-17T18:06:15Z) |
18:08 |
erle |
Wuzzy how did they get broken anyway? |
18:10 |
MinetestBot |
[git] JakobDev -> minetest/minetest: Add vcs-browser and contribute URLs to Appdata db9b3af https://github.com/minetest/minetest/commit/db9b3aff75d920585c793639050ccccb18972cca (2022-05-17T18:08:32Z) |
18:10 |
|
orwell96 joined #minetest |
18:11 |
alguien |
MTDiscord, are you sure? I've seen people who didn't die reaching 0 hp |
18:11 |
alguien |
and their health regenerated on its own |
18:11 |
MTDiscord |
<Warr1024> Which MTDiscord user are you talking to, IRC? |
18:11 |
alguien |
IhrFussel* |
18:12 |
MTDiscord |
<Warr1024> Ah, I see |
18:12 |
MinetestBot |
[git] savilli -> minetest/minetest: Remove confusing message in keybindings menu af37f9d https://github.com/minetest/minetest/commit/af37f9dc542e31f34d53da995ed86f2a3eacd9a5 (2022-05-17T18:11:08Z) |
18:12 |
MTDiscord |
<Warr1024> Many things cannot be reliably cheated from the client side, IF the server mod/game code is setup to actually properly enforce certain mechanics and not just rely on default client behavior. |
18:14 |
erle |
yeah |
18:14 |
erle |
like, autokill |
18:14 |
erle |
ridiculously easy to figure out if someone is hitting every entity in reach |
18:15 |
Wuzzy |
erle: see #12328. sfan5's fault, apparently |
18:15 |
ShadowBot |
https://github.com/minetest/minetest/issues/12328 -- DevTest: Fix broken PNG textures by Wuzzy2 |
18:15 |
Wuzzy |
erle: what do you mean with "bresenhams axe"? |
18:16 |
erle |
yeah i want to konw which tool breaks PNGs. sfan5 do you remember? |
18:16 |
erle |
Wuzzy the algorithm you use for tool usage calculation is similar to bresenhams line drawing algorithm. a friend of mine saw it and named yours “bresenhams axe”. |
18:17 |
Wuzzy |
ahhhh |
18:17 |
Wuzzy |
yeah you're not the first one saying this =) |
18:18 |
Wuzzy |
i think i still have the Lua version of the code lying around somewhere... |
18:18 |
erle |
Wuzzy if you can maek the lua code to determine usages left a mod it would be very nice |
18:18 |
erle |
for tooltips |
18:18 |
sfan5 |
erle: apparently I ran this: nice oxipng -o max games/devtest/mods/basetools/textures/basetools_* |
18:18 |
sfan5 |
but I think there was something else, maybe stripping the iCCP chunk? |
18:18 |
erle |
sfan5, oh is it better than optipng in some way? |
18:19 |
sfan5 |
it can multithread |
18:19 |
sfan5 |
also zopfli if you want it |
18:19 |
Wuzzy |
See #11110. It contains the old Lua code I used to test out my maths. it's not ready-to-use for your usecase tho |
18:19 |
ShadowBot |
https://github.com/minetest/minetest/issues/11110 -- Fix number of tool uses being off by 1..32767 by Wuzzy2 |
18:19 |
erle |
oh |
18:19 |
erle |
zopfli i see |
18:19 |
MTDiscord |
<Warr1024> I use advpng for zopfli ... and lately I've started using ect too, but it's super-brute-force slow |
18:19 |
Wuzzy |
i use zopflipng for all my compressions =) |
18:20 |
MTDiscord |
<Warr1024> multithreading is less important of a feature as long as you have more files to process than you have cores :-) |
18:20 |
erle |
a paletted map that my WIP mod generates. it will be used for treasure maps. so … water is nice, but how to draw land? https://mister-muffin.de/p/9Li5.png |
18:21 |
MTDiscord |
<Warr1024> erle I don't see anything wrong with your depiction of land per se as it is, unless you wanted to add like topographical contours or something. |
18:22 |
fluxionary |
what's the proper way to check whether 2 objects/entities are the same thing? |
18:23 |
alguien |
"<erle> ridiculously easy to figure out if someone is hitting every entity in reach" <- but not if they're doing it with a csm |
18:23 |
Wuzzy |
fun fact: zopfli is named after this tasty thing: https://en.wikipedia.org/wiki/Zopf |
18:24 |
sfan5 |
fluxionary: == works on ObjectRefs |
18:24 |
erle |
alguien the CSM hits even invisible entities ;) |
18:24 |
fluxionary |
sfan5: thanks :) |
18:24 |
erle |
Warr1024 yeah i want contours or anything. like, this is just too barren. |
18:24 |
sfan5 |
because there is only one ObjectRef instance in Lua for an object |
18:24 |
alguien |
erle, *typetype* not anymore |
18:24 |
erle |
Warr1024 like what if there is a chest hidden in a cave. |
18:27 |
alguien |
please tell me a hacked client cannot set node metadata? |
18:27 |
alguien |
in particular the "owner" field |
18:28 |
Krock |
they cannot set node metadata on their own |
18:29 |
MTDiscord |
<Warr1024> erle, instead of contours, one other option might be to detect certain terrain features like leaves/trees or something... |
18:29 |
Krock |
interaction with formspecs are needed to do that |
18:30 |
sfan5 |
the reason the node range is the way we discussed is that the client has limited knowledge of surroundings that it can't see, which is complicated to avoid |
18:30 |
sfan5 |
map modifications aren't like that because it's not up to the client but the server, and the server only does what it is programmed to |
18:31 |
Wuzzy |
erle: probably not exactly what you asked for, but this PR has some utility functions: #12047 |
18:31 |
ShadowBot |
https://github.com/minetest/minetest/issues/12047 -- Add helper functions to make tool usable n times by Wuzzy2 |
18:31 |
MTDiscord |
<Warr1024> Re: what hacked clients can do: generally they can move about freely, and trigger any event on the server they want, e.g. punch, place, dig. There may be some protections in place to ensure those inputs are "sane" given the circumstances, but to allow for the game to be playable under lag, there's a lot of leeway by default. What then CAN'T do is break out of the game logic once those events have been passed onto the server. The |
18:31 |
MTDiscord |
entity:on_punch or on_dig_node event or whatever is called, and then the game/mods have the chance to validate that action and the client can't do anything about that. |
18:32 |
MTDiscord |
<Warr1024> We could do a line-of-sight check for the nodes that CSMs are allowed to see, but (1) raycasts can be rather expensive, and (2) what counts as "sight" can be actually quite tricky. If you can see the corner of a node, but a raycast from your camera to the center of the node would be obscured, should that "count" or not, and if so, how?? |
18:33 |
MTDiscord |
<Warr1024> I've actually done a number of modding things where things are affected by "line of smell", so I'm doing like a floodfill algorithm to find things that influence something ... accounting for that kind of thing would make stuff get crazy complex pretty fast. |
18:34 |
alguien |
Krock, can you elaborate please? |
18:34 |
Pexin |
floodfill, you mean just shortest path? |
18:35 |
alguien |
Krock, if a node provides a form, they can do that? |
18:35 |
Krock |
alguien: only if the formspec is not properly secured |
18:35 |
Krock |
there's no direct way for clients to write to any metadata |
18:36 |
Krock |
difference being that clients can read out all public node metadata fields because they're needed for static formspec field expressions |
18:36 |
Krock |
i.e. ${metadata_key} unfolds to the value contained in the node metadata |
18:39 |
erle |
what hacked clients are usually used for is either combad (between hackers), which looks like dragonballZ basically |
18:39 |
MTDiscord |
<Warr1024> Pexin, yeah, floodfill as in connected to the starting point by any path up to a certain distance, but not in a particular direction (it can bend and twist around obstacles). Actually ... similar to how light spreads in MT already. |
18:39 |
alguien |
Krock, ay |
18:39 |
erle |
and also quality-of-life things, like automatic eating or killing specific mobs in range |
18:40 |
erle |
also tracers |
18:40 |
erle |
tracers are so nice |
18:40 |
erle |
should be in vanilla IMO |
18:40 |
erle |
but the vast majority of cheat client uses i have witnessed was neither. it was automatic building. |
18:40 |
|
sobkas joined #minetest |
18:40 |
MTDiscord |
<Warr1024> What hacked clients are good for is very often orthogonal to what I care about in my game so I usually don't try to block use of the clients preemptively, unless I see them actually find a way to meaningfully cheat, and actually try to exploit it... |
18:40 |
erle |
and automatic mining. |
18:41 |
alguien |
Krock, how would an insecure formspec look like? If you want, you can PM me btw |
18:41 |
MTDiscord |
<Warr1024> to be fair, alternative clients of various kinds (including "cheat" ones) can sometimes actually be necessary for certain users to some extent for accessibility reasons. |
18:41 |
erle |
yeah, i think non-vanilla minetest clients are only a problem if your game is … very … grindy. |
18:41 |
erle |
like, imagine if you could autodig, no one would complain about repixtures dig times. |
18:41 |
Krock |
alguien: assume the fields are only sent to players who can see the formspec, that's mistake M°1 |
18:41 |
Krock |
* N°1 |
18:42 |
erle |
CSMs are super cool |
18:42 |
Pexin |
CSMs wear sunglasses and drive motorcycles. It's true. |
18:42 |
Krock |
number two being that you might forget to restrict list interactions which would "not be possible" in theory |
18:43 |
alguien |
Krock, like "read only" fields? |
18:43 |
alguien |
(not sure if that's a thing in formspec, just guessing) |
18:43 |
Krock |
and final thought N°3 is that you assume a certain input value. dropdown selected values or table selections can be spoofed and be any invalid value |
18:44 |
Krock |
more like, actions that the player is not allowed to do with the permissions you give them. |
18:45 |
Pexin |
alguien: it's the difference between what the formspec can do, and how the formspec is triggered |
18:45 |
Pexin |
re number 1 |
18:46 |
alguien |
Krock, thanks for the explanation |
18:46 |
erle |
full recognition before processing, always |
18:50 |
Krock |
yw |
19:04 |
|
proller joined #minetest |
19:20 |
erle |
is there a function like minetest.find_nodes_in_area_under_air() that just finds any nodes under air except air nodes? |
19:21 |
erle |
(most air nodes are under air) |
19:22 |
Krock |
custom_homebrew_functio394_that_uses_vmanip |
19:22 |
Krock |
that's the function you'll need |
19:23 |
MTDiscord |
<Warr1024> that, or find_nodes_in_aread_under_air("group:everthing_but_air_group_i_hacked_in") |
19:24 |
MTDiscord |
<Warr1024> also are you sure you really want "under air"? Or maybe you really want "under airlike" or even "under non-sunglight_propagates) ... vmanip would seem like the way to go then. |
19:25 |
Krock |
group:not_in_creative_inventory,flower,tree,wood,cracky,stone,level,crumbly if you need a hacky temporary solution |
19:27 |
|
Sven_vB joined #minetest |
19:31 |
MTDiscord |
<Warr1024> heh, not bad, though not sustainable... |
19:31 |
MTDiscord |
<Warr1024> also, if those are MTG groups, they might not work quite right for a minclon* game. |
20:48 |
|
wallabra joined #minetest |
21:02 |
|
Sven_vB joined #minetest |
21:05 |
|
sometalgoo joined #minetest |
21:12 |
|
hlqkj_ joined #minetest |
21:22 |
|
harry-wood joined #minetest |
21:32 |
|
wallabra joined #minetest |
21:34 |
|
harry-wood joined #minetest |
21:59 |
|
harry-wood joined #minetest |
22:00 |
|
Sven_vB joined #minetest |
22:30 |
|
harry-wood joined #minetest |
22:35 |
|
panwolfram joined #minetest |
22:39 |
|
sometalgoo joined #minetest |
23:10 |
|
queria joined #minetest |
23:56 |
|
AliasAlreadyTake joined #minetest |
23:58 |
|
Lesha_Vel joined #minetest |
23:59 |
fluxionary |
has anyone ever proposed changing the fluid mechanics in minetest? i'm mostly interested in conversations people had about this in the past, if there were any. |