Time Nick Message 10:03 deltasquared is there a way to adjust the camera offsets for minetest's 3D modes? I'm currently tinkering with some DIY head-mounted displays and have been trying out sidebyside mode with it. 10:04 deltasquared for a 5" display panel it works quite well but at 7" the inter-pupil distance seems off. can't focus on it and keep seeing double. 10:04 deltasquared (one could argue why don't I move the panel further away, well, I wanted the increased FOV...) 10:27 ANAND deltasquared: player:set_eye_offset might be of help to you 10:34 deltasquared ANAND: I do believe that's just a singular camera. if I move it left/right it affects both views in a 3d mode. 11:26 ANAND Oh sorry, I misunderstood 12:10 jas_ hi 12:11 * jas_ is suddenly reminded of per-player fov overrides 12:11 jas_ that got merged, right? 12:12 jas_ i tried the android build of minetest, to see if 3d mode worked, and for my phone it did not. :( i wasn't super disappointed. if it did work, there could be a sub-game made specifically for VR mode. 12:12 jas_ game* 12:15 ANAND jas_: Yes, the FOV PR has been merged 12:15 deltasquared I think per-player FOV override was a bit of a mistake considering my use case above for instance... if the server can force change it, it would be somewhat disorienting for a player using a HMD to play the game. 12:15 ANAND (yay :D) 12:16 deltasquared in that same sense, being able to yank a player's head around suddenly is also somewhat questionable 12:16 deltasquared (even if only virtually) 12:16 deltasquared I may have to patch that out of my client if I proceed with this HMD experiment. 12:16 jas_ so that's a job for the mod, or game. that's why i suggested a VR-centric game 12:17 jas_ the server really mandates all the examples you say, so. if the game doesn't suddenly jerk the player, all should be well. 12:17 nepugia I don't think allowing such stuff to be set serverside is a good idea anyway 12:17 jas_ like, why would you set mouse sensitivity server-side? 12:17 jas_ why? you wouldn't. leave my FOV along! haha 12:18 deltasquared to be fair, it's always a balance 12:18 deltasquared I think if it's used to "scope in" on a weapon for instance, then fine 12:18 nepugia deltasquared, yes, but such stuff doesn't make sense serverside either 12:18 deltasquared however, there should be a user-side override for if the server decides to be a douche. 12:19 ANAND nepugia: I'd agree with you, if SSCSMs were a thing :) 12:19 nepugia alternatively, dont play on the server 12:19 nepugia ANAND, sscms? i use spcms for server provided, i wonder what the ss stands for :P 12:20 ANAND Server-sent 12:20 nepugia and since csm exist spcsm shouldnt be that far away 12:20 ANAND Means the same thing though :) 12:20 nepugia indeed 12:20 ANAND I really want to believe you when you say that it isn't far away :D 12:20 deltasquared I think SSCSMs are a bad idea for separate reasons... 12:20 deltasquared mostly to do with software freedom principles 12:21 ANAND But I'm afraid no core dev does 12:21 nepugia deltasquared, explain 12:21 deltasquared nepugia: it's like javascript. I should have the right to not execute code that I do not wish to. 12:21 nepugia well... you do in all instances, you control the code on your computer anyway 12:22 deltasquared what I'm saying is that servers ought to be prepared for clients to refuse to execute SSCSMs. 12:22 nepugia I wouldnt not implement spcsm just because one potential implementation could be shitty 12:22 jas_ but some games (and hence servers) will depend on the presence of certain CSMs 12:22 nepugia deltasquared, simple solution, kick the client :) 12:22 jas_ and if they're not found, the server can then kick the player. 12:22 nepugia some gamemodes will simply Not work without them 12:22 ANAND ^ 12:22 deltasquared if that is the case, then so be it, but the user needs to have that choice 12:23 nepugia sure 12:23 jas_ have a choice as to whether to play the game? 12:23 jas_ some aspects of any game are client-side 12:23 ANAND Clients would be allowed to opt out of it, there's been discussion regarding this, IIRC 12:23 nepugia some mods could work in a degraded mode i suppose 12:23 deltasquared right, I have nothing against the server saying "then nope, you're out" 12:23 nepugia still, more effort for modders :D 12:23 ANAND But if SSCSMs are critical to the gameplay, the server could just kick non-conforming clients, as jas_ said 12:23 deltasquared ANAND: I have no issue with this 12:23 jas_ and nepugia, simultaneously heh 12:24 nepugia I do need spcsm for some nicer UI anyway ... 12:24 deltasquared if that becomes an issue it becomes a problem the game can't solve anyway, so it's beyond the scope of the game 12:24 deltasquared (as in, mods requiring SSCSMs and the user objects, that's a design issue, can't fix that) 12:24 jas_ the game is comprised of the mods 12:24 jas_ why would the game not work with its mods? you're talking about a server operator? 12:25 jas_ because if a server operator adds a non-comforming mod to their game, then that's that 12:25 deltasquared that's not to say I object to SSCSMs entirely - they have awesome potential. but the choice must be present. 12:25 jas_ the choice is always present 12:25 jas_ don't connect to the server, if you don't trust it 12:26 jas_ but to be fair, there's like 500 options for disabling CSM 12:26 jas_ and precisely zero for sending scripts to the client. 12:26 jas_ talk about "should have a choice" hehe 12:29 nepugia Garrys mod was nice :'( 12:29 deltasquared nepugia: oh? did that have server side sent scripts? 12:29 nepugia Indeed it did 12:29 deltasquared once again don't get me wrong as a modder they have such awesome potential, not least being non-laggy custom tool logic and so forth 12:29 nepugia you could check SERVER and CLIENT to check on which side you were on, and terminate accordingly 12:30 nepugia I was about to implement a whole inventory system and looting system in gmod with a nice ui, and then those **** changed their EULA to a consumer-hostile bullshit license 12:31 deltasquared by "those ****" I assume you mean facepunch studios? 12:31 deltasquared (I don't live that far away from their offices...) 12:31 nepugia also never seen so many spelling mistakes in a EULA 12:31 nepugia deltasquared, :) 12:31 nepugia Just a bit angry about it, you know... 12:32 nepugia oh well, i reckon they will get problems with the GDPR with that EULA sooner or later 12:34 deltasquared software freedom does seem to be a thing that riles up people who care about it. for good reason of course. (I myself just now had trouble containing myself on the SSCSMs issue. they're a bit of a double edged sword.) 12:35 nepugia Well, they will have to be disabled-by-default at the start anyway, maybe use that as a setting? 12:35 deltasquared as in, they would be opt-in initially? 12:35 deltasquared first I heard of it... 12:35 nepugia gmod had a nice thing that specified how much content you wanted to download from servers (from the source engine). with "map only" "sound and textures" and "all content" 12:35 nepugia maybe minetest could do that similar, and possible with a per-server overide 12:36 nepugia deltasquared, well, that is how you implement features no? :) 12:36 nepugia on the develop branch i mean, i highly doubt that the first iteration will be that fun 12:36 deltasquared right, so you could locally control the scripts that executed? pick your own version and be satisfied it Respects Your Freedom™? 12:37 nepugia eh, i'd rather say execute this script or not, and then tell the server about it maybe? 12:37 nepugia rewriting scripts is infeasible to verify on the serverside, and some server people are really scared of people cheating 12:38 nepugia in any case, i do think that the boat here is different to javascript, because there it isnt at all required, here it kinda is, for some mods anyway 12:41 deltasquared I personally would be fine if there was an option to allow me to use a local version of the script, which, funnily enough, is basically existing CSMs. but it appears that the core devs aren't willing to trust pure CSMs to the same degree. 12:41 deltasquared which is BS, because that's the DRM fallacy. ultimately the client can send whatever it damn pleases, how it reached that conclusion is irrelevant 12:43 nepugia I coded my gmod server mod with an api and dont trust clientside code, yeah, but allowing arbitrary csms is a different design 12:43 nepugia it emphasises apis between server and client, rather than coherent mods in itself 12:43 nepugia Gmod had a setting to allow csms on the server, and a setting for the client to refuse server code 12:44 deltasquared the former seems a bit of a misfire. the client could always conceal the fact it runs CSMs. 12:44 deltasquared (another thing I've always thought of doing if it became an issue) 12:45 nepugia Well, yeah... but those are "evil hacked clients" you know? :) 12:45 nepugia and valve thinks they can find people using those with their VAC anti cheat 12:46 deltasquared there's another problem with SSCSMs as well... it forbids future alternate implementations. 12:46 nepugia I sure am happy that i dont get banned on mt for "cheating" especially since i can trigger the "nep moved too fast" thing if i want 12:46 deltasquared or at the very least significantly raises the barrier 12:46 nepugia deltasquared, it does not 12:46 nepugia that is just plain bs, js doesn't prevent new js from ebeing written either 12:46 deltasquared like what if I want my own client written? at that point, code in the client is indistinguishable from CSMs, because at that point I could make it do whatever I like 12:47 deltasquared (ok, now I am conflating server-sent and non-server-sent CSMs, but I still think my last point is important in itself) 12:47 nepugia deltasquared, there is certain network traffic the client generates... and then those that the mods generate 12:47 nepugia those are different, and are threated differently, even if you could generate all of them from C++ if you wanted 12:48 nepugia and do your "what if i", i say "so what" 12:48 deltasquared nepugia: again, if the server decides it will not serve me because of that, then again that is a problem that the game can't solve. that's a case of me complaining about my freedoms to the server op. 12:48 nepugia deltasquared, .... any? 12:49 nepugia i honestly dont see what you tried to respond to with that 12:49 nepugia if you dont like a server dont play on it 12:49 nepugia if you want to get code, /any/ code that makes the server what it is changed you have to complain any way 12:49 nepugia regardless of whether it is gameplay specific or you just dont like how the code looks 12:51 deltasquared no, it's a case of I'll play anyway and I'll disable what I don't like. if the server doesn't like *that* that's their problem. 12:51 nepugia Not really no 12:51 deltasquared assuming it's detectable for specific "I don't like X, that's turned off" 12:52 nepugia whether your client becomes non-conformant or not doesnt change what the server sends 12:52 nepugia and just plain kicking you for beeing non-conformant is an option 12:53 nepugia also, if there /ISNT/ a nice way to just specify "this client doesn't like mods and will plug its ears" then modders will go through great lengths to try and detect that at runtime with bullshit in luaspace 12:53 nepugia so to me, just specifying 'im out' is a lot better than trying anything else 12:53 deltasquared nepugia: hence what I said about being detectable. a client could always fake acceptance of an SSCSM; if it's really important for gameplay, the onus is then on the server component to kick the player if they don't recieve some kind of packet they expect. 12:53 nepugia deltasquared, then you just have to guard every network call, that sucks 12:53 nepugia and you kick people for your own broken code 12:53 nepugia in all cases, a bad solution 12:54 nepugia so don't pretend to accept it if you dont 12:54 deltasquared nepugia: you should be doing that anyway! don't trust the network. 12:54 nepugia you just make the game worse for everybody :) 12:54 nepugia deltasquared, no, freenode shouldnt ban me because i suspended the network 12:54 nepugia laptop* 12:54 nepugia they should timeout... gracefully 12:54 nepugia you know, like minetest already does? 12:54 deltasquared I meant guard every network call, not the kicking part 12:55 nepugia you don't need to guard them really, unless you mean for bad data 12:55 nepugia but that isnt modders domain anyway 12:55 deltasquared more the general case of not trusting clients. 12:55 nepugia that should have been done by the engine a long time ago 12:55 nepugia Yes... but that doesn't have anything to do with server provided mods... 12:56 nepugia honestly at this point i'd just say replicate exactly what gmod did and call it a day :P 12:56 nepugia (well, not exactly, but that system worked just fine) 12:57 deltasquared I think what I'm trying to get across, is that a) people may well do my suggestions above and b) not for nefarious reasons either, but rather because they value their software freedom. 12:58 nepugia No, intentionally lying about your intentions isnt about software freedom 12:58 nepugia If you dont like it simply dont execute it 12:58 nepugia (and be kicked for it, if it is essential to have) 12:59 jas_ you can do tons of stuff with lua modding to manage connections/clients/players 12:59 jas_ plus server settings 12:59 jas_ and manage the settings via lua too, which is also fun 13:00 jas_ i dunno what you're trying to restrict, but i bet you can do it 13:00 jas_ my favorite setting is `enable_strict_protocol_checking' or some such, i don't use it tho 13:00 jas_ i use disable_anticheat = true, but i still see NoCheat stuff in the log, so i guess that's a separate thing 13:01 jas_ and can act on it in minetest.register_on_cheat 13:01 deltasquared nepugia: hmm. I guess at the end of it a mod "really" needing an SSCMS is again a design or possibly political problem. so I guess perhaps saying NAK to the server is better then. (assuming the choice is there, which it really should be.) 13:01 nepugia deltasquared, just tell the server that "okay, then im out" 13:01 jas_ mod's don't need server sent CSMs, the server does. 13:02 deltasquared nepugia: I'd rather it be "I'm not going to do that, what are you gonna do about it" 13:02 nepugia mark mods that /neeed/ clientside code as essentiall for that purpose and done 13:02 nepugia deltasquared, why? it's plain capability checking 13:02 nepugia if your client does not support the execution of essential clientside mods then it shouldnt connect 13:03 jas_ if the server wants the client to run a script, it has to tell them the URL where to get it. there's no option for the convienence 13:03 nepugia jas_, you what? 13:03 jas_ the client user may want to click ok to download the script, instead of having to follow a url 13:03 nepugia what are you even talking about 13:03 nepugia minetest doesnt use https as a transport 13:03 deltasquared jas_: I'm fairly certain SSCSMs would be sent the same way as current mod resources. 13:03 deltasquared nepugia: so then it'd need to be either "this server has no CSMs", "this server has optional CSMs", or "this server has mandatory CSMs and gameplay will break otherwise". 13:03 nepugia or /anything/ that has a valid URI scheme 13:03 jas_ currently, if you - as a server - want a client to run a particular CSM, you have to direct them to it, and they have to download it 13:04 nepugia deltasquared, yes 13:04 nepugia jas_, no 13:04 nepugia jas_, you just send it to them... 13:04 jas_ SSCSM is sending the script automatically from the server to the client. this is for convienence 13:04 jas_ you cannot 13:04 nepugia yes, one can 13:04 jas_ because this debate rages on 13:04 deltasquared nepugia: that is the current situation. he's talking about what we have right now. 13:04 deltasquared not what we're going towards with SSCSMs 13:04 jas_ and the ability for the server to send scripts to the client for the client to execute simply does not exist 13:04 jas_ at least, not in this reality 13:04 nepugia We were talking about the situation that will be the entire time, why change the topic now... 13:05 jas_ sounds right in line to me 13:05 deltasquared nepugia: don't look at me, maybe they think it'll never happen? 13:05 jas_ my point is, it's not a matter of paranoia. 13:05 jas_ it's a matter of convenience. 13:05 nepugia there exists /NO/ way for csm and ssm to communicate at this time, so they mostly are pointless 13:05 jas_ yes there is 13:05 jas_ mod channels 13:05 deltasquared actually I was working on that 13:05 jas_ !mod sneakjump csm 13:05 MinetestBot jas_: There are no results for this query :( 13:05 deltasquared (not mod channels, they suck) 13:05 jas_ hold on i made a mod for this 13:05 jas_ so? 13:05 jas_ they still qualify 13:05 jas_ you said /NO/ way 13:05 nepugia not really 13:06 jas_ oh 13:06 jas_ ok 13:06 nepugia There is no way to specify that i need to talk to a clientside mod for the purposes of prediction and stuff 13:06 nepugia for example, that certain code gets called /clientside/ for on_dig 13:06 jas_ https://forum.minetest.net/viewtopic.php?f=53&t=22683 13:06 jas_ !title 13:06 MinetestBot jas_: [csm] sneakjump 100 - Minetest Forums 13:06 jas_ then there's a link to the SSM there, and they do indeed communicate 13:07 nepugia and no, calling on_dig serverside to mod channels does not count, that just triples the latency (while in reality latency prevention was the goal here) 13:07 jas_ it would be /nice/ if i could send the client script to the client, from the server 13:07 jas_ that's all i'm saying, adding to the conversation 13:07 jas_ thank you very much 13:07 nepugia That is only one tiny aspect of spcsm, and indeed a hard dependency 13:10 deltasquared on a completely tangential topic, I wonder how far one could go with SSCMSs if they were restricted to something essentially like how bpf is used in the linux kernel right now... 13:10 deltasquared provably halting and can't breach the host program because everything is statically checkable. 13:11 deltasquared more of a security thing than a software freedom one in that case. 13:11 deltasquared you'd be limited to hooks of the type "action happens -> produce response" though. 13:12 nepugia Make a non-turing complete language if you want that i suppose 13:12 nepugia i don't know bpf, i only know pf and that is a firewall 13:12 deltasquared nepugia: bpf is more like a vm bytecode, so a language would be easier to write I suspect :) 13:12 nepugia deltasquared, "cant break the host programm" is not something you can verify 13:13 nepugia breach* 13:13 deltasquared nepugia: not for turing complete programs, no. the kernel can do it for bpf so there is precedent. 13:13 nepugia the kernel cant do it there either 13:13 deltasquared well nobody has managed to break it so far... 13:13 nepugia it can check it for /some/ cases 13:14 nepugia but if anything all those nice spectre exploits show you that hardware may not behave in the way the code suggests/or intends to 13:14 deltasquared I mean sure there will always be a bug at some point. I was getting at the fact that for normal programs it's intractable; for a specially designed non-turing-complete machine, it's merely hard 13:14 nepugia non-turing complete stuff for some point would be interesting, for example exactly for prediction of node placement and such 13:15 deltasquared oh yeah, for certain. it's the stuff it'd be good at. 13:15 deltasquared hmmmm... I wonder if it'd be possible to generate an array containing mesh triangles from such a hook 13:15 deltasquared (though that gets into trouble with loop checking I guess) 13:17 deltasquared there's always the "easy" way of enforcing loop bounds. inject a timeout into the loop condition so that past N loops it is forced to stop with a failure return value. 13:17 deltasquared getting into abstract CS here... 13:18 nepugia i don't mind :P 13:18 nepugia anyhow, nice discussion, i will try to get some time into my haiku patch now 13:20 deltasquared two things that always get me riled up... having to execute arbitary code someone else sent, and assuming that was already yielded, the potential security/sandboxing implications. :) 13:20 nephele (but am still here, but from haiku yaaay) 13:20 nephele security is hard 13:20 nephele sandboxing is hard 13:21 deltasquared aye, good job I don't have a day job doing that. I'd probably have thrown something by now. 13:21 nephele Just throw linux namespaces at it and pray" they are kernel devs, surely that is secure" 13:21 deltasquared unless it's user namespaces 13:21 deltasquared "where is your god now" 13:22 nephele try to derail any security discussion into a "why minix3 > linux" discussion or a "fuchsia > linux" discussion :P 13:22 deltasquared bah, linux > any of those on hardware support grounds and therefore actually getting used. >:D 13:23 nephele But but haiku runs on /my/ computer :) 13:23 nephele surely it will be the bestest soon(tm)... like with hardware acceleration :D 13:24 nephele Although... nothing but linux runs on my laptop, and still provides me with backlight control :( 13:24 nephele With errm, sensitive eyes that is a bit of a problem :) 13:28 deltasquared to solve my original question from earlier about inter-pupil distance, I'm considering buying a purpose built HMD instead which can be adjusted... 13:29 deltasquared turns out you can get the htc vive HMD alone (no controllers etc) for about 100 bucks or so 13:29 deltasquared as I don't have neither the playspace nor games supporting tracking input, this is not a problem for me :) 13:34 nephele the valve index can adjust the pupil-distance 13:34 nephele with a hardware dial i think 13:36 deltasquared nephele: it's newer though, so I wouldn't have much look getting just the HMD unit. 13:36 deltasquared *luck 13:37 deltasquared like discounted. 13:37 nephele You can get just the unit sure, but discounted i do doubt it too :D 13:37 nephele (why would anyone want to sell valve hardware? :P /s) 13:39 Ingar maybe the Epic store will get hardware and give it away for free 13:41 nephele lol 13:41 deltasquared nephele: according to the ifixit teardown, the htc vive's headset also is a mechanical adjust 13:41 deltasquared just a threaded rod with a knob. simples 13:41 nephele I mean, it has to be mechanical at some point :D 13:42 deltasquared aye, I was checking it wasn't some dumb software driven system is all 13:42 deltasquared seeing as I'll be using it on linux with pretty much nothing aware the display is special 13:43 deltasquared (well, bar minetest in sidebyside mode...) 13:43 nephele I mean, at some point it has to have a motor to do the physical moving :) 13:43 deltasquared nephele: no motors in the vive, all manual adjust 13:43 nephele can always try to hijack that 13:44 deltasquared including one for adjusting the distance the lenses are from the eyes, which is neat 13:44 nephele deltasquared: ah, well if it was software you could still hijack motors :D 13:44 deltasquared nephele: I'd rather not hack it if need be, I'd rather it just works as dumb hardware for now. 13:44 nephele yeah 13:45 CWz_ hey deltasquared 13:45 deltasquared seems according to thit at least http://doc-ok.org/?p=1763 , it just appears as a regular display to xrandr for instance. 13:45 deltasquared (that example uses nvidia binary drivers but the principle should be the same.) 13:45 nephele it is just a regular display :/ 13:45 deltasquared nephele: well actually it's two, just made to look like one 13:45 deltasquared I needed to check there wasn't weirdness there too 13:46 nephele X11 does that too, so whatever ;) 13:46 nephele Probably better that the software there does it rather than X11, i always get terrible overlaps in X11 (like window decorations on the other monitor and stuff) 13:46 MinetestBot 02[git] 04programmerjake -> 03minetest/minetest: Fix unwanted detaching when damage = 0 137603215 https://git.io/JeCqA (152019-10-05T13:45:05Z) 13:46 deltasquared nephele: that's one desktop across two screens, not two physical oled panels made to look like one screen. :P 13:47 deltasquared that's a point... the display is oled. I will have to be very careful 13:47 deltasquared no chat or HUD elements at all 13:47 deltasquared else I'll get burn-in o.o 13:47 deltasquared rip me 13:47 nephele You worry too much about burn in :D 13:47 deltasquared still, I'd rather not push it 13:47 nephele but if you want, just make it wobble a bit if you move around 13:47 nephele valves index doesn't have oled 13:47 deltasquared to be fair I find the hotbar and chat to look distracting in 3d mode 13:48 nephele i do like oled though 13:48 deltasquared they always appear at a fixed distance so it looks wrong 13:48 nephele deltasquared: probably want something like a hotbar attached to your hand or something 13:48 nephele vr stuff yay 13:48 deltasquared nephele: that's what minecraft vr does IIRC. 13:48 nephele thats what most vr games do 13:48 deltasquared nephele: yeah, but I was thinking of the closest example. :P 13:48 MinetestBot 02[git] 04Warr1024 -> 03minetest/minetest: Merge pull request #8166 from Warr1024/fix7020 13ad96df1 https://git.io/JeCqp (152019-10-05T13:46:57Z) 13:48 nephele but a HUD in a "you are in a helmet" sort of fashion would be fun 13:48 deltasquared (comparing minecraft to minetest, I have committed sin father) 13:49 nephele yeah, one is good, and the other is gpl 13:49 nephele :P 13:49 deltasquared careful, you'll get hit by thunder for that! xD 13:50 nephele I wanted to read up on BMessage anyway 13:53 deltasquared nephele: ... ah, they're samsung amoled panels! ok, that makes me relax a little 13:53 deltasquared I've had my s5 for ages with an amoled panel and it's never had burn-in, despite constant persistent UI elements from android 13:54 deltasquared it's about smartphone density as well. ~440dpi. 13:58 nephele I liked my S5 13:58 nephele sadly i destroyed it, oh well 13:59 nephele Personally i dont want a htc vive because the pixel density is too little for my taste 13:59 nephele tried it at a demo once 13:59 nephele asked whether it was the low-res dev kit, was told that it was the consumer one :/ 14:00 deltasquared nephele: it's denser than the current display I'm using in a google cardboard compatible headset. that one only gives me 400x480 per eye, and I can't even see all of it. 14:01 nephele that it is i suppose 14:01 nephele but for me either way a reason not to have a vr headset atm 14:01 nephele i do hope that valve index will be to my satisfaction, but i would want to try that before buying it too :) 14:03 deltasquared what do you think of the index controllers? assuming linux would at least be able to read the input data, I think that idea is dope. 14:03 deltasquared I'd like to know how exactly it senses fingers. 14:03 nephele I like the design, wanted to buy one of those 8 years ago... there were similar ones before 14:04 nephele but those looked more like borg hands than the sneaker valve one with laser tracking, and didnt have any sense of pressure 14:04 nephele I lost the link to time though :( 14:05 nephele (also lost a link about how to properly do multitouch form like 15 years ago, and that is much better than current multitouch... but that video had like 300 views so finding it again will be difficult... :( ) 15:15 * deltasquared puts in the order for the htc vive HMD 15:15 deltasquared fingers crossed... 15:16 VanessaE fun/ 15:16 deltasquared hey VanessaE 15:17 VanessaE hi 15:38 jluc ’ 15:38 jluc je pige pas : 15:38 jluc le filedset de saisie de l'url d'un forum est en display:non 15:39 jluc mais ça saisit qd mm une valeur (saisie, bien avant, dans un autre formulaire) 15:39 jluc je pense que c'est l'autocomplétion des formulaires par un plugin chrome 15:39 jluc mais dans la console 15:40 jluc document.getElementById('url_site') 15:40 calcul0n heu, t'es sur d'être sur le bon chan ? :) 15:41 jluc ah non zut 15:41 jluc dsl 15:46 Krock jluc: il y a un channel français -> #minetest-fr 16:17 MinetestBot 02[git] 04sfan5 -> 03minetest/minetest: Fix warnings in guiButton.h 13dde0628 https://git.io/JeCOf (152019-10-05T16:15:26Z) 16:32 nepugia Is there a technical reason why textures only allow a restricted charset in their names, and have to have a file extension? 16:36 Calinou nepugia: maybe so Irrlicht can detect their type? I don't know 16:36 Calinou (also, why is this an issue?) 16:37 Calinou it's considered good practice to make files have an extension and restrict their names to a sane charset for compatibility 16:37 nephele that might explain the second restriction but not the first :) 16:37 nephele compatibility with what? MS-DOS? 16:37 Calinou many, many applications which don't handle spaces and special characters well 16:37 Calinou including on Linux 16:37 Calinou even uppercase characters can be troublesome if you're not strict, due to Windows case sensitivity 16:37 nephele If apps don 16:38 nephele 't support utf-8 they should be fixed :/ 16:38 nephele Linux also has case sensitivity if you want 16:38 nephele but uppercase chars arent banned 16:38 nephele but instead for example : is banned, 16:39 Calinou I meant case insensitivity 16:39 nephele which i discovered only after having trouble for like half an hour to figure out why my textures werent working 16:39 nephele Calinou: yeah, ext4 has that too, optionally 16:39 Calinou ":" isn't allowed on Windows, which is probably why 16:39 Calinou that bit me a few times already 16:39 nephele it isn't? 16:39 Calinou yeah 16:39 nephele I thought only \ wasn't allowed huh 16:40 sfan5 no that's on linux 16:40 nephele no, / is on linux 16:40 sfan5 well you get what I mean 16:40 Calinou https://stackoverflow.com/questions/1976007/what-characters-are-forbidden-in-windows-and-linux-directory-names 16:40 sfan5 windows has problems with many more, namely: * . " / \ [ ] : ; | , 16:41 sfan5 though I'm not sure [] should be in that list(?) 16:41 nephele Wouldn't it make most sense to ban solely the offending characters? 16:41 nephele Also file extensions are annoying because i may want to have my textures be .png files, but someones texture pack may choose another format 16:42 nephele And windows supports files without extensions just fine, so i dont really see any issues there for compatibility 16:42 Calinou I think texture packs detect alternate extensions automatically, or I thought they did 16:42 sfan5 ^ 16:42 Calinou well, the issue is that on Windows, it will ask you which application should be used to open the file *every time* 16:42 Calinou it's not MIME type-aware, unlike other OSes out there 16:43 nephele maybe macos? linux usually just goes searching for magic :P 16:44 nephele I wonder if dofile() refuses to read files that are invalid on windows too 16:44 rubenwardy If you miss out the extension in minetest, it detects it automatically 16:45 nephele rubenwardy: that isnt my file name though, that to me is a wierd inconsistency 16:45 rubenwardy Lua does the same thing 16:45 rubenwardy As does python 16:45 rubenwardy Inferring extensions 16:45 nephele what? 16:46 nephele It just opens a different file? 16:46 rubenwardy require("a") could be a.lua or a/init.lua 16:48 nephele eh, that seams wierd.... 16:48 nephele i would expect that to open a file called a (assuming that require opens files) 16:51 `ani` this is shit? 16:52 Krock no 16:52 Krock this is Patrick 16:53 Krock relevant http://images.wikia.com/cardfight/images/8/85/No-This-Patrick-Meme-Template.jpg 16:53 * nepugia digs out this is patrick sparta remix 17:26 MinetestBot 02[git] 04SmallJoker -> 03minetest/minetest_game: Use background9 for backwards compatibility 131c413ed https://git.io/JeC33 (152019-10-05T15:55:42Z) 20:57 kurtzmusch are there any attempts at making underliquids nodes? 20:57 kurtzmusch i was browsing through the issues tagged 'feature request' and couldnt find any 20:57 kurtzmusch was there any attempt at doing this? 20:58 kurtzmusch example: a stair node without an air bubble 21:02 calcul0n the v5 has underwater plants which seem to do that 21:02 calcul0n no idea how they are implemented though 21:08 kurtzmusch in a hacky way 21:15 Pie-jacker875 hey would it be useful for anyone here to know that I get like <10fps on my RX 5700 XT 21:16 Pie-jacker875 performance was fine on my R9 390 22:25 hisforever Hi I have my own internet connection now. for a screenshot in Mt. is there a way to haide the inventory slots? 22:26 kurtzmusch F1 22:26 kurtzmusch or some F key 22:26 kurtzmusch i dont remebmer wich 22:27 hisforever Kurtzmusch: thanks 22:28 hisforever it was f1 my friend