| Time |
Nick |
Message |
| 00:17 |
|
chilledfrogs joined #minetest |
| 00:34 |
|
TheSilentLink joined #minetest |
| 01:08 |
|
smk joined #minetest |
| 01:18 |
|
sys4 joined #minetest |
| 01:26 |
|
liceDibrarian joined #minetest |
| 01:27 |
|
sys4 joined #minetest |
| 02:31 |
|
sparky4 joined #minetest |
| 02:58 |
|
ywang joined #minetest |
| 03:13 |
|
v-rob joined #minetest |
| 03:31 |
|
SoniEx2 joined #minetest |
| 03:52 |
|
YuGiOhJCJ joined #minetest |
| 04:14 |
|
Edgy2 joined #minetest |
| 04:36 |
|
SpaceManiac joined #minetest |
| 04:37 |
|
hook54321 joined #minetest |
| 04:46 |
MTDiscord |
<jordan4ibanez> Maybe I can ramble my questions into here |
| 04:46 |
MTDiscord |
<jordan4ibanez> Can we offset an entity's mesh? |
| 04:46 |
MTDiscord |
<jordan4ibanez> Entities aren't thread safe and don't plan on being thread safe right? |
| 05:00 |
|
MTDiscord joined #minetest |
| 05:07 |
|
muurkha joined #minetest |
| 05:19 |
|
doseijin- joined #minetest |
| 05:23 |
|
jaca122 joined #minetest |
| 06:03 |
|
xBarkDog joined #minetest |
| 07:02 |
|
s20 joined #minetest |
| 07:12 |
|
v-rob joined #minetest |
| 07:48 |
|
TheSilentLink joined #minetest |
| 08:23 |
|
calcul0n joined #minetest |
| 08:33 |
|
Evil joined #minetest |
| 08:54 |
|
SoniEx2 joined #minetest |
| 08:55 |
|
mrkubax10 joined #minetest |
| 09:15 |
|
clavi joined #minetest |
| 09:15 |
|
clavi joined #minetest |
| 09:26 |
mrkubax10 |
Is there any significant network performance difference between MineClone2 and Mineclonia? |
| 09:28 |
erle |
mrkubax10 i have not tested this, but i have worked in the past with cora (mineclonia) on improving network performance, while on the other hand ancient_mariner (mineclone 2) said he considered network stuff too difficult. |
| 09:29 |
erle |
mrkubax10 i believe the main user-facing difference right now is mineclone 2 simply not working on any devices due to filling the entire RAM with garba^H^H^H^H^H^H background music |
| 09:29 |
erle |
s/any/some/ |
| 09:29 |
erle |
mrkubax10 hold on i may have a script to estimate network stuff |
| 09:30 |
erle |
mrkubax10 this will analyze a minetest log with at least INFO log level https://git.minetest.land/Mineclonia/Mineclonia/src/branch/master/tools/analyze-packet-spam |
| 09:32 |
erle |
btw this is what i mean |
| 09:32 |
erle |
Sep 06 18:47:34 <[MTMatrix]><ancientmariner> I tried wireshark and my eyes glazed over. Network is a weakness in my skill set. Wouldn't know where to start |
| 09:33 |
erle |
mrkubax10 if you want to have a fun time, patch your client to output location of received packages that contain coordinates outside of your visual range. you may or may not find some coordinate exploits in mineclone2 that way ;) |
| 09:34 |
mrkubax10 |
lol |
| 09:34 |
mrkubax10 |
I will look into that then because I'm running MineClone2 server and it tends to be quite laggy sometimes |
| 09:41 |
erle |
mrkubax10 while i am not part of the current mineclonia effort, it would be a surprise for me if it somehow caused more network traffic. the developers are using cheat clients that are AFAIK very good at picking up stray network traffic. |
| 09:41 |
erle |
mrkubax10 though you should clarify what kind of lag you are experiencing |
| 09:42 |
erle |
mrkubax10 there can be server-side lag caused by excessive processing (e.g. taking a photo with my xcam mod does tens of thousands of raycasts currently and hangs the server for seconds) or it may be network. but network only becomes relevant when it is A LOT. |
| 09:43 |
erle |
mrkubax10 anyway, throw some logfiles into my script and i think it may be able to tell you things about the network level spam |
| 09:44 |
erle |
mrkubax10 so what kind of lag oo you have anyway? |
| 09:46 |
|
Fusl joined #minetest |
| 09:54 |
mrkubax10 |
players hanging midair for example, that happens quite often |
| 09:55 |
mrkubax10 |
same for mobs, items etc |
| 09:55 |
erle |
to everyone at the same time or a single player? |
| 09:55 |
erle |
if everyone at the same time: this is the server not keeping up with processing (probably) |
| 09:55 |
erle |
if a single player: network lag (probably) |
| 09:55 |
mrkubax10 |
ah so probably it's the server then |
| 09:56 |
erle |
this could still be better in mineclonia. i know cora did for example rewrite the entire fire system (though i think that ended up both in mineclone2 and mineclonia) to make it less laggy. |
| 09:56 |
erle |
try it out then |
| 09:57 |
erle |
also if you find coordinate exploits, keep them. they are funny. |
| 09:57 |
erle |
:3 |
| 09:57 |
erle |
(i am joking here: griefers will use them to destroy everyone's bases.) |
| 09:58 |
mrkubax10 |
lol |
| 10:11 |
|
appguru joined #minetest |
| 10:38 |
|
sfan5 joined #minetest |
| 10:56 |
|
Sobinec joined #minetest |
| 11:22 |
erle |
AFAIK the particles from running effects were used by a griefer to find people's bases on a server once |
| 11:39 |
|
cow321 joined #minetest |
| 11:45 |
|
syl joined #minetest |
| 12:00 |
|
Ingar joined #minetest |
| 12:43 |
|
parimek joined #minetest |
| 12:50 |
|
ShadowBot joined #minetest |
| 12:52 |
|
lemonzest joined #minetest |
| 12:54 |
|
kamdard joined #minetest |
| 12:55 |
|
imi joined #minetest |
| 13:08 |
|
MisterE123 joined #minetest |
| 13:32 |
|
bigfoot547 joined #minetest |
| 13:38 |
|
definitelya joined #minetest |
| 13:54 |
|
bwarden joined #minetest |
| 14:06 |
|
Izaya joined #minetest |
| 14:13 |
|
jaca122 joined #minetest |
| 14:29 |
|
MinetestBot joined #minetest |
| 14:30 |
|
illwieckz joined #minetest |
| 14:35 |
|
wsor4035 joined #minetest |
| 14:51 |
|
mrkubax10 joined #minetest |
| 14:59 |
|
ROllerozxa joined #minetest |
| 15:04 |
|
basxto joined #minetest |
| 15:07 |
|
jonadab joined #minetest |
| 15:30 |
|
e1z0 joined #minetest |
| 16:08 |
|
CRISPR joined #minetest |
| 16:22 |
|
JerryXiao joined #minetest |
| 16:42 |
|
iamweasel joined #minetest |
| 16:45 |
|
gera joined #minetest |
| 16:56 |
|
sparky4 joined #minetest |
| 17:03 |
|
mrkubax10 joined #minetest |
| 17:06 |
jonadab |
erle: background music feels like the sort of thing individual players ought to be able to configure away on the client side. |
| 17:06 |
jonadab |
(I mean, they can already change the volume...) |
| 17:07 |
erle |
jonadab yes, but that is a complaint you should give to mineclone2 and i doubt they will listen |
| 17:08 |
erle |
(given “my phone crashes with recent versions when loading the game” did not make them listen …) |
| 17:13 |
jonadab |
I mean, I have no real interest in Mineclone. Monster farms aren't really my thing. |
| 17:13 |
MTDiscord |
<warr1024> If the background music is just a shuffle of songs that occasionally play independently of what's going on in the game, then just distribute those as a separate OST and let people run their favorite media player. If they're actually integrated with events that happen in the game, then it's easier to justify putting them in MT. |
| 17:13 |
sfan5 |
!tell Wuzzy do you have a good suggestion how I should translate 'schematic' for WorldEdit into German? |
| 17:13 |
MinetestBot |
sfan5: I'll pass that on when Wuzzy is around |
| 17:14 |
erle |
Warr1024 i suggested to just have a separate mod, but the “vision” trumps good gamery i guess |
| 17:16 |
jonadab |
Separate mod doesn't really put the power in the hands of individual players anyway. |
| 17:16 |
jonadab |
That's still the server admin deciding for everyone. |
| 17:16 |
erle |
true |
| 17:16 |
jonadab |
Which is fine for things like which blocks and recipes should exist. |
| 17:16 |
jonadab |
But seems odd for background music. |
| 17:16 |
erle |
yeah i find the hardcoded music on servers really annoying |
| 17:17 |
erle |
but “annoying” is one level below “won't run on some devices because entire RAM is filled with decoded wav files” (that one should be fixed in 5.8.0 btw) |
| 17:17 |
erle |
i mean decoded wav samples, it's ogg vorbis files obv |
| 17:20 |
erle |
the schematic thing is interesting. apparently many people do not translate it at all, even though i think “bauplan” fits really nice. i wonder why. |
| 17:21 |
jonadab |
A lot of languages use loan words for technical terms. |
| 17:21 |
jonadab |
I don't know anything about this specific case, though. That's just in general. |
| 17:22 |
erle |
this reminds me of being in the s-bahn with a british lady recently and both of us being amused that “bitte beachten sie beim ausstieg den abstand zwischen zug und bahnsteigkante” is the german version of “please mind the gap” |
| 17:24 |
|
darylsun joined #minetest |
| 17:30 |
jonadab |
I can almost imagine the German public actually taking the time to read a sign that long. |
| 17:30 |
jonadab |
Whereas, "please mind the gap" is about four times as much text as a typical American will read. |
| 17:33 |
erle |
it is not a sign |
| 17:33 |
erle |
it is the thing they say when there is an actual gap while the train is approaching the station |
| 17:33 |
jonadab |
Oh, an announcement on an intercom or something? |
| 17:33 |
erle |
but they say it in german and then english, creating this hilarious contrast |
| 17:33 |
erle |
yep |
| 17:33 |
jonadab |
I see. |
| 17:34 |
jonadab |
I have very little experience with passenger rail. Most of the times I've ridden on a train, it was in an amusement park. |
| 17:34 |
erle |
are you in a place where they do not have trains? |
| 17:34 |
jonadab |
We have _freight_ trains. |
| 17:34 |
erle |
i chatted with a girl from morocco, they have only one train line or so |
| 17:34 |
erle |
oh, are you from US? |
| 17:34 |
jonadab |
Yes, I live in Ohio. |
| 17:35 |
MTDiscord |
<warr1024> tbh if you want to put in game-driven background music and want players to have a choice on the music, and are okay with not having an "default" music (and thus ship it as a separate mod), you can just ship the game with empty music files (like zero-duration silent ogg files) and then the music media itself can be separate mod(s). |
| 17:35 |
|
v-rob joined #minetest |
| 17:35 |
erle |
yeah so in europe rail is big, s-bahn is urban transit stuff for a big city (like hamburg or berlin) and surrounding suburbs, often with a third rail and partially underground at least. |
| 17:35 |
jonadab |
Passenger rail used to be a common thing here, but there are not very many people left alive who remember it at this point. After WWII, everybody could afford to get a car, and that was the end of THAT. |
| 17:36 |
MTDiscord |
<warr1024> so you can still hook up the necessary game events and have the server-side code do the playing, and then players can get the music via CDB for SP, and for servers you can leave the music blank server-side and use the poorly-documented "sound packs" feature of MT client-side. |
| 17:36 |
erle |
i think there was also stuff like the big streetcar conspiracy |
| 17:36 |
erle |
Warr1024 oh tell me more about sound packs |
| 17:36 |
jonadab |
There are some big cities in North America that have significant passenger rail systems. The best is probably Chicago's. |
| 17:36 |
|
Izaya left #minetest |
| 17:36 |
jonadab |
But I live in a city of ten thousand people. |
| 17:37 |
MTDiscord |
<warr1024> I barely know about sound packs. Apparently there's like a "sounds" dir in MT that you can install sounds to, sort of like the way the "textures" works for texturepacks, and MT will just use those instead. I don't think there's a mechanics for having multiple of them or switching them out other than just adding/removing files or something, though. |
| 17:37 |
MTDiscord |
<warr1024> I've never personally used it but heard from IIRC multiple sources that they ARE a thing, even if the workflow is half-baked. |
| 17:38 |
|
Izaya joined #minetest |
| 17:38 |
jonadab |
For switching, you could probably just make the directory a symlink. |
| 17:38 |
jonadab |
Minetest wouldn't notice unless it goes out of its way to check for that, and there's no reason for it to do so. |
| 17:39 |
ROllerozxa |
the concept of "sound packs" in minetest is interesting because it means you can point a singleplayer game's sounds folder into it and bypass media loading altogether |
| 17:39 |
MTDiscord |
<warr1024> Could be. Not having a UI for it (let alone proper package management) though still makes it inaccessible for a big hunk of the playerbase. |
| 17:41 |
|
jaca122 joined #minetest |
| 17:41 |
MTDiscord |
<warr1024> That's an interesting point ... though the fact that the media cache is naively persistent (i.e. if you don't take measures to purge it yourself) it seems like just automatically scraping installed packages for media to prefill the cache would accomplish roughly the same thing. |
| 17:41 |
jonadab |
Sure, but how much overlap is there between "players who don't know what a symlink is" and "players who want to customize things like the music"? I'm not sure about game music specifically, but my _overall_ impression is that most non-technically-inclined users pretty much never customize anything ever. |
| 17:42 |
jonadab |
Although, hmm. |
| 17:42 |
jonadab |
I'm forgetting how much more opaque MS Windows is about things like symlinks. I retract my previous statement. |
| 17:43 |
MTDiscord |
<warr1024> The original discussion wasn't about players wanting to customize music, it was about players having a way to opt in/out of music at all. |
| 17:43 |
jonadab |
True. |
| 17:43 |
jonadab |
Which they _should_ be able to do, client-side. |
| 17:43 |
muurkha |
why would there be any correlation at all between technical knowledge and *wanting* to customize things? |
| 17:44 |
jonadab |
muurkha: I'm not sure, but there is. |
| 17:44 |
muurkha |
I don't think there is |
| 17:44 |
Ingar |
muurkha: if you want to customize things, you need to figure out how to do it |
| 17:44 |
Ingar |
if you don't figure out the knowledge, then you won't be customizing |
| 17:44 |
muurkha |
exactly. I think non-technically-inclined users pretty much never customize anything because they can't figure out how, not because they don't want to |
| 17:45 |
MTDiscord |
<warr1024> People who don't have technical knowledge often don't actually realize that there's something they could want. And those who only idly want to customize, but know they don't have the skills to do so, are also less likely to voice those wants. Which is just as well I guess, because we have our hands full already with the people who could customize things if there were a way, but there isn't... |
| 17:45 |
muurkha |
maybe wiser people have a wider variety of desires |
| 17:46 |
muurkha |
people who are ignorant of shaders probably won't want to edit shaders, though they might want to make visual effects |
| 17:46 |
MTDiscord |
<warr1024> wiser people also tend to pick up a lot of skills in the process of becoming wise, including technical ones, so it's still pretty moot. |
| 17:47 |
MTDiscord |
<warr1024> and people who just have general wisdom also have a better foundation to get those skills. |
| 17:47 |
jonadab |
I mean, I work with the public a fair amount, and my observation has been, that a lot of people don't customize even extremely basic stuff like the _text of their own resume_ any more than they absolutely have to. (They get a template resume, change the bare minimum they have to change, mostly their name and their last three employers' names and addresses and phone numbers, and leave everything else |
| 17:47 |
jonadab |
exactly as it is in the example.) |
| 17:47 |
Ingar |
for a lot of people, modding a game by placing a file in thge game directory is too complicated |
| 17:47 |
MTDiscord |
<warr1024> Haha, just because they don't try to customize also doesn't mean they won't complain that things aren't already to their liking 😆 |
| 17:48 |
|
Fulgen joined #minetest |
| 17:48 |
muurkha |
jonadab: maybe they don't like their resume because they associate it with stress |
| 17:48 |
MTDiscord |
<warr1024> Ingar: it's not just "complicated" that can make it impractical. Mobile tends to try to hide the fact that a filesystem exists from most users. |
| 17:49 |
Ingar |
mobile/console is a whole different can of worms ofc |
| 17:49 |
jonadab |
^ |
| 17:49 |
Ingar |
those systems were design "don't touch" to begin with |
| 17:49 |
Ingar |
*designed |
| 17:49 |
jonadab |
Mostly. |
| 17:49 |
jonadab |
They let you change like six things. |
| 17:50 |
|
fgaz_ joined #minetest |
| 17:50 |
erle |
the actual original complaint of mine was about mineclone2 authors putting style over substance (i.e. valuing “game has music” over “game starts on all devices that it previously started on”) |
| 17:51 |
erle |
i mean, almost any other project would revert a change if it turned out it accidentally drops support for a wide range of devices |
| 17:51 |
MTDiscord |
<warr1024> It depends on how important style IS to the game. |
| 17:51 |
erle |
and not go like “but artistic vision! the engine is at fault!” |
| 17:52 |
MTDiscord |
<warr1024> I would agree that Mineclone is not the kind of game that should be focused on like cinematic shit at the expense of functional shit, as it seems to me very much intended to be for sandboxy multiplayer collaborative projects with a wide audience. |
| 17:52 |
erle |
jonadab the funniest customization story is that people kept complaining to jwz that xscreensaver is not theme-able and so he pointed out it actually is, but no one ever sent him any themes for like 17 years even though he asked them to. |
| 17:53 |
MTDiscord |
<warr1024> but every game does also have to pick how narrow of an audience it focuses on. For example, NodeCore reaches a much broader audience in terms of technical capabilities (much more efficient and less demanding on hardware) but narrow cognitively. |
| 17:53 |
erle |
Warr1024 this kind of stuff fixes itself in the long-term. i have looked at both mineclone2 and new mineclonia issue trackers and the people are friendly and more constructive in the latter, the game supports more hardware configurations, has more features and less code and does not deliberately break stuff. so ig the future is there, eventually. |
| 17:54 |
erle |
almost nothing in mineclone2/mineclonia necessitates a chonky supercomputer, except mapgen ig |
| 17:54 |
MTDiscord |
<warr1024> how does mineclonia handle music? Is there just none and people are encouraged to make mods if they want to support it ingame? |
| 17:55 |
erle |
or some clown spawning 3000 particle effects obv but for that reason dragonfire can turn particle rendering off ig |
| 17:55 |
erle |
there are in-game jukeboxes and i think that was it |
| 17:56 |
MTDiscord |
<warr1024> I've mostly avoided music where there isn't some clear integration with the game. NodeCore has no music. For Piranesi Redo, I'm actually adding a buttload of music, but there are tons of musical cues that it's all tied to, e.g. area-specific music, or music on special events; access to new music is also sort of part of the "reward" for finding new areas. |
| 17:57 |
erle |
well, keep in mind that this currently fills the RAM with decoded tracks |
| 17:57 |
MTDiscord |
<warr1024> Does it actually like decode the whole ogg to PCM at load time? That would be amazingly shitty. |
| 17:57 |
erle |
so maybe you might want to have some dynamic media approach here if not all of them are immediately available |
| 17:58 |
MTDiscord |
<warr1024> I don't know how much dynamic media would help if players should be able to get into a state where they eventually would have them all loaded anyway, and I don't want to force players to quit and rejoin just to clear cache shit. |
| 17:58 |
erle |
i can not confirm this right now (i have no time), but reportedly that is the reason why adding a shitload of music crashes minetest with mineclone2. |
| 17:58 |
erle |
and it is fixed in 5.8.0 |
| 17:58 |
erle |
but some people (*cough*) are too eager to add shit that does not work instead of waiting for a release |
| 18:00 |
MTDiscord |
<warr1024> Okay, there's a possibility I may be pushing to drop 5.7.0 support ASAP anyway, so maybe that's not a problem. I'll just make one final 5.7-compatible release before merging in the music stuff. I was debating whether to support 5.7 with the upcoming music or not but if 5.7 runs into RAM issues, that's another reason not to. |
| 18:00 |
MTDiscord |
<warr1024> A nice thing about CDB is that users on old versions of the engine can continue to get old versions of the packages as long as you're careful about updating your compatibility info for each release. |
| 18:01 |
MTDiscord |
<warr1024> Haha, I get the not wanting to wait for release thing though, it's maddening. And often people don't know that something doesn't work until they've had enough testers on it to catch a rare problem ... and I bet for MineClone2's playerbase, running out of RAM is not super common and especially not amount devs. |
| 18:02 |
erle |
Warr1024 the problem is that the “old version” already crashes. but given the ”fast and loose” style that mcl2 is developed (like WAY worse than the ”fast and loose” style that minetest has), this is to be expected. |
| 18:02 |
erle |
like with minetest you have devs that actually listen to “this is a regression” |
| 18:03 |
erle |
Warr1024 btw, if you have any in-game scenes where the filtering working *correctly* is visible in minetest in nodecore, please tell. like which nodes etc. |
| 18:03 |
erle |
i want to get this fix right and i worry that there may have been more breakage i have not found. |
| 18:04 |
MTDiscord |
<warr1024> I mean it's okay to make a mistake from time to time, like merging a feature that you THOUGHT was tested and working only to discover that it doesn't work for a certain segment of your audience. At that point though, you have to decide whether you're going to find a workaround for some people (it has to be something that's practical enough for them to actually do), back out the change, or just kick a bunch of people out of your audience |
| 18:04 |
MTDiscord |
(which is a shitty thing to do with no warning or options). |
| 18:04 |
erle |
yeah |
| 18:04 |
ROllerozxa |
when you're talking about mineclone OOMing low-mem devices, are you talking like, old 768MB RAM android devices? |
| 18:04 |
erle |
i am talking about like smartphones |
| 18:04 |
erle |
2GB or so? no idea. |
| 18:05 |
erle |
but like NO justification for not supporting those just over adding some music |
| 18:05 |
muurkha |
it makes a difference if they're 16MiB smartphones or 16GiB smartphones |
| 18:05 |
muurkha |
depending on your value system |
| 18:05 |
erle |
look, i don't have a smartphone to run minetest on |
| 18:05 |
MTDiscord |
<warr1024> The sanest thing to do, to me, if I were running a project like MC2, would be to just back it out, say "oops sorry" and then either hold it on a side branch until I EOL pre-5.8 versions, or work on a workaround or other compromise. Assuming 5.8 is actually coming as soon as it's supposed to, and actually fixes the issue, it shouldn't even be necessary to consider booting out some audience members. |
| 18:06 |
erle |
i just think not backing out and saying “okay, music is an optional add-on” is valuing things differently than me and user-hostile especially to poor people (who might not be able to afford a new phone to play more minetest) |
| 18:06 |
MTDiscord |
<warr1024> It's not that there's NO justification for it, it's just that whatever justification they have is almost certainly not justification we would agree with. But a project CAN have weird priorities even if those don't make sense to us. |
| 18:06 |
erle |
Warr1024 the thing is, almost everything i have heard about your dev efforts sounds like “the sanest thing to do”. you have high standards (and it shows in the quality of your work). |
| 18:06 |
muurkha |
poor people mostly have better smartphones |
| 18:07 |
jonadab |
In principle there is also the option of saying "Users with not much RAM should probably avoid upgrading to the new version of the game, until they have the new version of the engine." |
| 18:07 |
muurkha |
both because they can't afford laptops and because they don't have the luxury of a sedentary daily routine that would make the laptop useful |
| 18:07 |
erle |
muurkha i don't mean “poor = bad hardware”. i mean “poor = can not easily acquire new hardware if support is dropped because that costs money” |
| 18:07 |
muurkha |
"new" is gradable |
| 18:07 |
ROllerozxa |
2GB sounds plenty even if android eats up a percentage of it, I thought we were talking about a 768MB-mem device like my nexus 7 tablet where you're gonna quickly hit your head in the ceiling when trying to do anything |
| 18:08 |
erle |
jonadab cdb actually does not offer that option |
| 18:08 |
MTDiscord |
<warr1024> Personally I'd just back it out and say "music is not optional, it's just not gonna happen until 5.8 at least". Whether it later becomes optional might depend on the options I have as a developer. If I really want having music to be the default (removing it might dramatically change the tone of the game, and users very often do not change defaults) then I just at least need to make it technically work. |
| 18:08 |
erle |
ROllerozxa mineclone2 is the only game that hits this limit AFAIK. 2 have 2GB RAM on my thinkpad and everything minetest that hits some limit hits some GPU limit (or CPU limit) much earlier before RAM becomes a factor. |
| 18:08 |
jonadab |
erle: Public servers don't either, I suppose. ALthough the server admins could hold off up upgrades that are problematic. |
| 18:09 |
erle |
jonadab the simple tip for server admins is “run mineclonia instead, it is almost the same game, but made with people who were alienated from the original because of the way it is developed” |
| 18:10 |
erle |
and also it contains less exploitable bugs, but finding those is left as an excersize for the reader ;) |
| 18:10 |
erle |
fewer |
| 18:10 |
erle |
not less |
| 18:11 |
muurkha |
'fewer' is less ambiguous in this case but 'less' is also correct |
| 18:11 |
jonadab |
"fewer" in this context means "not as many"; whereas "less" means "not as exploitable". |
| 18:12 |
jonadab |
So both are grammatically valid, but they mean different things. |
| 18:12 |
erle |
fewer |
| 18:12 |
erle |
you can see the reason in the mcl2 bug tracker |
| 18:12 |
muurkha |
https://en.wiktionary.org/wiki/less#Usage_notes |
| 18:12 |
erle |
there are a few very funny discussions about bugs |
| 18:13 |
ROllerozxa |
I mean it would be interesting to consider converting textures to use compressed pixel formats (e.g. ETC* or DXT) to reduce the memory (GPU or RAM?) usage, but the functionality for that in IrrlichtMt is currently under threat of extinction |
| 18:13 |
erle |
where the question gets asked “should this actually be fixed?” |
| 18:13 |
erle |
which is funny |
| 18:13 |
erle |
because the discussions are not only extremely uninformed at times but also take more effort than fixing it |
| 18:13 |
erle |
ROllerozxa no that is not interesting, it is totally inappropriate for pixel art |
| 18:13 |
muurkha |
jonadab: > Nevertheless, less has been widely understood and commonly used as a synonym for fewer since it first appeared in Old English as læs. |
| 18:14 |
jonadab |
muurkha: That depends on context. |
| 18:14 |
erle |
ROllerozxa also texture memory has so far not been an issue GPU-wise AFAIK, the ridiculous number of drawcalls (e.g. inventory) has been |
| 18:14 |
ROllerozxa |
erle: I'm obviously not talking about pixel art |
| 18:15 |
jonadab |
muurkha: If he'd just said "less bugs", it would mean the same thing as "fewer bugs". But that isn't the sentence. |
| 18:15 |
erle |
ROllerozxa i think stuff that is not used gets deleted, even if it is useful, so make your case soon ig :D |
| 18:16 |
* jonadab |
idly wonders whether it's possible to embed steganographic messages in MT textures. |
| 18:17 |
erle |
jonadab use tga_encoder and assign the same color to different palette indexes. boom! no visual difference, but a hidden message! |
| 18:18 |
jonadab |
erle: And a customized client could retrieve the message? |
| 18:18 |
jonadab |
Yeah, I guess it seems obvious that it could. |
| 18:18 |
erle |
no, because all textures are converted to 32bpp RGBA AFAIK |
| 18:18 |
jonadab |
Oh. |
| 18:18 |
erle |
for display |
| 18:18 |
jonadab |
And that's done server-side? |
| 18:18 |
erle |
what? |
| 18:19 |
jonadab |
I'm not asking if most users would see the message. The whole point is they wouldn't. |
| 18:19 |
erle |
the conversion is done in the client |
| 18:19 |
erle |
but the image would be in the cache ig |
| 18:19 |
jonadab |
I'm asking if a client who expects the message to be there could use a customized client to retrieve it. |
| 18:19 |
erle |
yes, if you make your client handle decoding of images different |
| 18:19 |
erle |
ly |
| 18:20 |
jonadab |
So e.g. a MT server could be used as a covert distribution mechanism for private messages, to get past censorship or something. |
| 18:20 |
jonadab |
It was just an idle thought. |
| 18:21 |
erle |
don't do it, stuff is not encrypted |
| 18:21 |
erle |
but there exists a chat CSM that does actual encryption for chats |
| 18:21 |
erle |
end to end |
| 18:22 |
erle |
it should be in the waspsaliva repository somewhere |
| 18:22 |
|
Talkless joined #minetest |
| 18:22 |
jonadab |
I don't actually have a personal use for such a technology at this time. It was just something that occurred to me randomly, as things do. |
| 18:23 |
erle |
i know that i used to communicate with other players by means of written books or signs |
| 18:23 |
erle |
so the server owner could not easily spy on us and we could not slip up in chat |
| 18:23 |
erle |
but censorship, nah |
| 18:25 |
erle |
jonadab if you can actually program, i'd suggest to figure out a way to make minetest protocol-level encrypted or at least authenticated properly. |
| 18:25 |
erle |
(like, no-MITM-possible at least) |
| 18:30 |
|
sometalgoo joined #minetest |
| 18:34 |
|
ghoti joined #minetest |
| 18:35 |
muurkha |
jonadab: the context does make it ambiguous |
| 18:40 |
MTDiscord |
<warr1024> MT could definitely serve as a covert communication channel in cases where you have an otherwise secure network and can trust a server operator better than you trust other parties around you. If we had e2e encryption for chat and other communications though, it could make a big difference; as it is, the circumstances that would make it actually viable are very hard to evaluate safely. |
| 18:42 |
erle |
Warr1024 e2e encryption is why i want more love for CSM (e.g. CSM on CDB) and not more of those “SSCSM” ideas that are high-in-the-sky ideas |
| 18:42 |
erle |
well that and a lot of debugging aids (e.g. scripting players to test games) |
| 18:45 |
MTDiscord |
<warr1024> e2ee feels a tad heavy for CSM, i.e. a pure lua implementation or something. Seems like something that MT should seriously consider pulling in a new dependency for (we SHOULD be able to find a small enough one). I mean, I'm not against a pure lua CSM for it, if the API is capable, but I have my doubts about how well it will perform, and it might be tricky to get enough API support to handle establishing trust, verifying identities, and |
| 18:45 |
MTDiscord |
keeping track of ratcheting keys and such safely. |
| 18:46 |
|
appguru joined #minetest |
| 18:51 |
erle |
Warr1024 wdym “too heavy” |
| 18:51 |
erle |
oh you mean the details i see |
| 18:52 |
erle |
i have my doubts about the minetest API being, uh, appropriate. given the zstd and png encoding options. |
| 18:52 |
erle |
you need some library where you can do absolutely nothing wrong when creating a binding and i know no such libary |
| 18:53 |
erle |
especially with crypto |
| 18:53 |
erle |
anyway, i gotta go |
| 18:53 |
MTDiscord |
<warr1024> Haha, I was only talking about the basic design of the thing ... if you have to realistically factor in the ability of a bunch of vaguely disinterested hobbyist developers doing a thing right that somebody might eventually stake their safety on, well, yeah, that's a whole nother level. |
| 18:54 |
erle |
it's not the ”vaguely disinterested”, it is the tendency to “do the simplest thing that works somewhat” |
| 18:54 |
erle |
which is okay-but-disappointing for non-security things |
| 18:54 |
erle |
but pointless for security stuff |
| 18:54 |
MTDiscord |
<warr1024> I wouldn't want to lull somebody into thinking that they're definitely safer with any e2ee we implement than otherwise, but I do want people whose only options are "whatever minetest can do" and "totally fucked" to be able to at least try something. |
| 18:54 |
|
v-rob joined #minetest |
| 18:54 |
erle |
unless we are talking hash based signatures or HMAC |
| 18:54 |
erle |
those are simple and work |
| 18:55 |
erle |
i mean look at those JWT libraries that allowed the ”null” signature algorithm (the signature algorithm used is part of the message, i.e. under control of an attacker) |
| 18:59 |
MTDiscord |
<warr1024> If you make security tools expose all the complexities of their inner workings to consumers and force them to understand what they're doing, then you pass the potential for mistakes on downstream and end up with a ton of broken implementations everywhere. If you try to constrain things so that all mistakes are precluded, then you end up making an API that's so restrictive that it misses a ton of peoples' use-cases and people feel |
| 18:59 |
MTDiscord |
compelled to reimplement things downstream, and you end up with a ton of broken implementations everywhere. |
| 19:01 |
erle |
Warr1024 i do not think you have to expose everything. i think you have to get it right. |
| 19:01 |
erle |
Warr1024 and there are ways to get it wrong by providing only one pre-set, if that preset is a bit stupid. |
| 19:02 |
erle |
Warr1024 for example, if you have something that supports both signatures and MACs, then having a function validate(message, signature_or_mac) would be simple and stupid |
| 19:03 |
erle |
(why that is dangerously bad API design is left as the exercize for the reader, because it is a funny joke IMO) |
| 19:07 |
|
Krock joined #minetest |
| 19:07 |
mrkubax10 |
What is the oldnode in after_destruct? |
| 19:08 |
MTDiscord |
<warr1024> the node that was just destroyed (and is therefore not there anymore) |
| 19:09 |
|
TheSilentLink_ joined #minetest |
| 19:09 |
mrkubax10 |
yeah I know lol but what type is this, just a position? |
| 19:10 |
|
gxt joined #minetest |
| 19:13 |
BuckarooBanzai |
it is of type `{ name="some:nodename" param2=0 }` iirc |
| 19:15 |
|
sparky4 joined #minetest |
| 19:23 |
mrkubax10 |
ah alright |
| 19:29 |
|
sparky4 joined #minetest |
| 19:45 |
|
sparky4 joined #minetest |
| 19:58 |
|
kamdard joined #minetest |
| 20:14 |
|
krupa joined #minetest |
| 20:14 |
krupa |
hi |
| 20:16 |
|
krupa joined #minetest |
| 20:16 |
krupa |
hello |
| 20:27 |
|
krupa joined #minetest |
| 20:28 |
* krupa |
is now playing: sorohanro - Etnographia (ID: 57924) |
| 20:29 |
|
appguru joined #minetest |
| 20:30 |
|
krupa joined #minetest |
| 20:36 |
|
Leopold joined #minetest |
| 20:48 |
|
fluxionary joined #minetest |
| 21:07 |
|
sparky4 joined #minetest |
| 21:11 |
|
copygirl joined #minetest |
| 21:38 |
|
staceee joined #minetest |
| 22:27 |
|
TheSilentLink_ joined #minetest |
| 22:34 |
|
TheSilentLink joined #minetest |
| 22:40 |
|
TheSilentLink joined #minetest |
| 22:52 |
|
appguru joined #minetest |
| 23:01 |
|
Sokomine joined #minetest |
| 23:14 |
|
z812 joined #minetest |
| 23:20 |
|
v-rob joined #minetest |
| 23:32 |
|
panwolfram joined #minetest |
| 23:34 |
|
TheCoffeMaker joined #minetest |
| 23:38 |
|
sys4 joined #minetest |
| 23:44 |
|
sys4 joined #minetest |
| 23:50 |
|
srifqi joined #minetest |
